By Lindsay Gehman and Saachi S. Gorinstein

GenAI is changing how brands and marketing agencies create, edit, distribute, and scale influencer content. GenAI tools can:

• Draft and edit captions, scripts, and talking points.
• Generate or modify images, video, and audio.
• Clone voices and create digital replicas of real people.
• Create virtual influencers with no corresponding human identity.
• Produce consumer reviews and testimonials at scale.

These capabilities raise a recurring question for brands and agencies: when must consumers be told that AI was used to create or modify influencer content? There is currently no general FTC rule requiring disclosure solely because AI was used. However, existing consumer protection law applies regardless of whether content is created by a human, an AI system, or both, and AI-specific disclosure obligations are emerging in state law, platform policies, and the EU AI Act. Under federal law, the more useful question is whether the use of AI changes what consumers are likely to believe about the advertisement, particularly its source, its authenticity, or the experience of the person making the endorsement.

The answer depends heavily on how AI is used. This article examines three categories:

• AI-assisted real influencers. A real influencer remains the speaker and source of the endorsement, while GenAI helps create or modify the content (see AI-Assisted Real Influencers).
• AI replicas of real influencers. AI creates content depicting an identifiable real person saying or doing something the person did not actually record, such as through voice cloning, digital replicas, or synthetic video (see AI Replicas of Real Influencers).
• Fully synthetic influencers. AI creates the influencer itself, resulting in a fictional persona that may have a realistic appearance, voice, personality, and social media presence without any corresponding human identity (see Fully Synthetic Influencers).

Each category raises different transparency concerns, but the analysis begins with familiar advertising law: what the advertisement communicates to a reasonable consumer, and whether that communication is truthful and not misleading. This article then offers a framework for assessing AI transparency across all three categories (see Assessing AI Transparency) and practical strategies for managing risk (see Managing Risk).

Key Rules at a Glance

The legal framework for AI-generated influencer content is discussed in detail in GenAI in Advertising and Influencer Marketing: The Legal Framework (2026 Update). In summary, the authorities that most often drive the analysis are:

• FTC Act Section 5. An advertisement is deceptive if it is likely to mislead reasonable consumers about something material, judged by its net impression. Objective claims must be substantiated before they are made, including claims suggested or drafted by AI (15 U.S.C. § 45(a)(1); FTC Policy Statement on Deception, appended to Cliffdale Assocs., Inc., 103 F.T.C. 110, 174 (1984)).
• Endorsement Guides. Endorsements must reflect the honest opinions and experience of the endorser, an endorser who is represented as using a product must be a bona fide user, and material connections must be disclosed clearly and conspicuously. The Guides apply to virtual influencers, and advertisers, endorsers, and intermediaries such as agencies may all be liable (16 C.F.R. §§ 255.0(b), 255.1(a)–(f), 255.5).
• Consumer Reviews and Testimonials Rule. Businesses may not create or disseminate reviews or testimonials that misrepresent that the reviewer exists, used the product, or had the experience described, and violations may result in civil penalties (16 C.F.R. § 465.2).
• State AI disclosure laws. New York requires advertisements that include AI-generated synthetic performers to disclose that fact (N.Y. Gen. Bus. Law § 396-b), and California prohibits using undisclosed bots to mislead people about their artificial identity in order to incentivize a purchase (Cal. Bus. & Prof. Code §§ 17940 to 17943).
• Rights of publicity and digital replica laws. State laws restrict unauthorized commercial use of a person’s name, likeness, or voice, and several states limit the enforceability of vague digital replica provisions in contracts (see, for example, Cal. Lab. Code § 927; N.Y. Gen. Oblig. Law § 5-302).
• Platform policies and the EU AI Act. Major platforms require labels for realistic AI-generated or altered content, and the EU AI Act requires disclosure of deepfake content (Article 50(4), Regulation (EU) 2024/1689).

AI-Assisted Real Influencers

AI assistance involving real influencers is likely the most common category. A real influencer remains the source of the endorsement, while GenAI may be used to:

• Draft, rewrite, or translate captions and scripts.
• Generate alternative versions of an approved post.
• Retouch photographs or video, or generate or modify backgrounds.
• Modify the influencer’s appearance or voice.
• Generate portions of a video.
• Develop product claims or talking points.

Legal Risks

There is no general FTC rule requiring an influencer to disclose AI use merely because AI helped draft or edit content. However, these uses vary substantially in their legal significance.

Key risks include:

• Undisclosed material connections. A paid relationship remains a material connection that generally must be clearly and conspicuously disclosed, whether or not AI was used to create the post (16 C.F.R. § 255.5(a)). Disclosing the sponsorship does not cure a false or misleading claim in the advertisement (16 C.F.R. § 255.1(a); FTC, .com Disclosures: How to Make Effective Disclosures in Digital Advertising (Mar. 2013)).
• AI-generated claims. GenAI can produce confident statements that are unsupported or incorrect. For example, an AI tool asked to turn approved talking points into a script may add that “clinical research proves” the product improves a condition by 40% when no such research exists. That a claim originated with AI does not reduce the advertiser’s substantiation obligation. The same risk arises when an influencer runs approved claims through a GenAI tool to put them in the influencer’s “own voice,” which can introduce new numbers, studies, or comparisons that were never approved. Influencers who make unsupported claims that go beyond their personal experience may also be liable (16 C.F.R. § 255.1(e)).
• Altered product demonstrations. An AI edit that adjusts lighting presents a different issue than an edit that makes a skincare product appear to eliminate wrinkles that remain visible in the original footage. The question is whether the modification creates a materially different impression about the influencer, the product, or its results. Long before GenAI, the Supreme Court upheld the FTC’s position that an undisclosed simulated demonstration offered as visual proof of a product claim is deceptive (FTC v. Colgate-Palmolive Co., 380 U.S. 374 (1965) (undisclosed mock-up in a shaving cream commercial)). The UK Advertising Standards Authority has applied similar reasoning to beauty filters in influencer advertising (ASA Rulings on Skinny Tan Ltd and We Are Luxe Ltd t/a Tanologist Tan (Feb. 3, 2021)).
• Altered appearance. The Endorsement Guides treat as deceptive an image that misrepresents a material attribute of the endorser, such as using a photograph that exaggerates the results the endorser achieved (16 C.F.R. § 255.1(g)). AI edits to an influencer’s appearance raise the same concern when the alteration relates to the product’s claimed effect.
• Platform labeling. Realistic AI alterations may trigger platform labeling requirements even where the law does not. Platform policies typically distinguish between realistic synthetic content, which must be labeled, and productivity uses such as drafting scripts or captions, which generally need not be.

AI Replicas of Real Influencers

Voice cloning, generative video, and digital replicas can create content depicting an influencer saying or doing things the influencer never recorded. The resulting advertisement may be difficult for consumers to distinguish from genuine content. For example, a brand might generate a video in which an influencer appears to say, “I’ve used this product every day for six months, and it completely transformed my skin,” although the influencer never recorded the statement or used the product for six months.

Legal Risks

The central FTC concern is not that the content is artificial, but whether consumers are likely to attribute synthetic statements, conduct, opinions, or experiences to the real person depicted. Key issues include:

• Two distinct disclosure questions. A sponsorship disclosure, such as #ad, tells consumers about the commercial relationship. It does not tell consumers that the influencer never recorded the video, spoke the words, or had the experience portrayed.
• False attribution. An endorsement must reflect the endorser’s honest opinions and experience, and an advertiser may not use an endorsement to convey a representation that would be deceptive if made directly (16 C.F.R. § 255.1(a)). An endorsement may not be reworded in a way that distorts the endorser’s opinion or experience (16 C.F.R. § 255.1(b)), and if the advertisement represents that the influencer uses the product, the influencer must have been a bona fide user (16 C.F.R. § 255.1(c)). AI makes false attribution more convincing but does not change the underlying deception principle.
• Statements after the relationship changes. A replica can continue generating content long after the original campaign. An advertiser may use a celebrity endorsement only while it has good reason to believe the endorser continues to hold the views presented, and may continue representing that an endorser uses a product only while it has good reason to believe the endorser remains a bona fide user (16 C.F.R. § 255.1(b), (c)). A replica that keeps praising a product after the influencer has stopped using it, or after the relationship has ended, raises both endorsement and publicity concerns.
• Disclosure of synthetic content. There is no general FTC rule requiring every synthetic image, voice, or video used in advertising to carry an AI label, and no FTC rule squarely addresses the authorized use of a real influencer’s replica. The FTC’s proposal to extend its Impersonation Rule to individuals has not been finalized and would principally address unauthorized impersonation. The risk is greater where a reasonable consumer would believe that the real influencer personally recorded the content, made the statement, demonstrated the product, or experienced the claimed result.
• Consent is not substantiation. An influencer’s approval of synthetic content cannot convert an experience that never occurred into a genuine first-person endorsement.
• Publicity and replica laws. Unauthorized replicas may violate state right of publicity and digital replica laws and support Lanham Act false endorsement claims (see, for example, Tenn. Code Ann. § 47-25-1101 et seq.; 15 U.S.C. § 1125(a)(1)(A); Waits v. Frito-Lay, Inc., 978 F.2d 1093 (9th Cir. 1992)). State publicity claims have proven viable against AI voice clones (Lehrman v. Lovo, Inc., No. 1:23-cv-08269, 2025 WL 1902547 (S.D.N.Y. July 10, 2025)). Even where the influencer has signed an agreement, vaguely drafted replica provisions may be unenforceable in some states (see Cal. Lab. Code § 927; N.Y. Gen. Oblig. Law § 5-302; 815 ILCS 550/1 et seq.). A general grant of rights to use an influencer’s name, image, likeness, voice, and content may not clearly extend to a replica that generates new statements the influencer never made.
• EU disclosure. Replica content in campaigns that reach the EU may require deepfake disclosure under the EU AI Act (Article 50(4), Regulation (EU) 2024/1689).

Fully Synthetic Influencers

A virtual influencer may have a realistic appearance, name, voice, personality, social media presence, and apparent life story without corresponding to any real person. The persona may post, interact with followers, express opinions, and promote products much like a human creator. The more realistic the persona, the less obvious its synthetic nature may be, particularly when it speaks in the first person about apparent personal experiences.

Legal Risks

Virtual influencers raise two distinct deception questions:

• Identity. Whether consumers understand that the influencer is synthetic rather than a real person.
• Experience. Whether the advertisement represents or implies that an actual product experience occurred when it did not.

For example, a virtual influencer post stating, “I’ve been using this product for six months, and my skin has never looked better. #ad,” discloses the sponsorship. Even if the persona is clearly identified as AI-generated, neither disclosure creates an actual six-month product experience.

Key issues include:

• Fabricated experiences. The Consumer Reviews and Testimonials Rule prohibits testimonials that materially misrepresent the testimonialist’s existence, use of the product, or experience (16 C.F.R. § 465.2). Brands should not use AI to create fictional customer reviews or use a synthetic persona to convey a first-person experience in a way that suggests a real person had it. A useful test is whether a human influencer who never used the product could truthfully make the same statement. If not, a synthetic influencer should not make it either.
• Reported experience and summaries. A synthetic persona that reports what actual customers say, rather than claiming its own experience, shifts the analysis from the endorser’s honesty to substantiation. The underlying reviews must be genuine and representative, and the presentation must not distort what consumers think of the product (16 C.F.R. § 255.2(d)). Brands that use AI to summarize verified reviews should retain the underlying data and a record of how the summary was produced.
• Material connections. Virtual influencers can be endorsers under the Endorsement Guides, so material connection disclosure obligations apply (16 C.F.R. §§ 255.0(b), 255.5).
• Identity disclosure. There is no FTC rule requiring every virtual influencer to be labeled as AI-generated, but omitting that information may be more significant where an advertisement depends on consumers believing the persona is human. New York now requires advertisements that include a synthetic performer to disclose that fact conspicuously, where the advertiser has actual knowledge of the synthetic performer’s use (N.Y. Gen. Bus. Law § 396-b (effective June 9, 2026)). A photorealistic virtual influencer in a campaign reaching New York audiences is likely to fall within the law. California’s BOT Act may apply where the persona’s interactions with followers are automated, though perhaps not where human operators write the posts (Cal. Bus. & Prof. Code §§ 17940 to 17943).
• Placement of disclosures. A disclosure that appears only in a profile or bio may not shape the overall impression of a post that users encounter through recommendations, sharing, or search. The Endorsement Guides state that a material connection disclosure placed only on a profile page is not clear and conspicuous, and similar reasoning applies to AI disclosures (16 C.F.R. § 255.0(g)(9); FTC, Disclosures 101 for Social Media Influencers (Nov. 2019); FTC, .com Disclosures: How to Make Effective Disclosures in Digital Advertising (Mar. 2013)).
• Resemblance to real people. A persona described as fictional can still implicate a real person’s rights if its face, voice, or mannerisms evoke an identifiable individual. Courts have long held that deliberately imitating a distinctive voice or evoking a celebrity’s identity to sell a product can violate the right of publicity, even without using the person’s actual image or recording (see, for example, Midler v. Ford Motor Co., 849 F.2d 460 (9th Cir. 1988)). Prompts instructing an AI tool to create a persona that resembles a particular person may later serve as evidence of intent.
• Substantiation. Objective product claims made by a synthetic influencer remain advertising claims that require substantiation.

Assessing AI Transparency

A rule requiring disclosure whenever AI is used does not fit the range of technologies in influencer marketing. A more useful analysis considers whether AI:

• Creates or simulates a real person’s statements or experience.
• Creates a fictional person or persona.
• Would lead a reasonable consumer to believe that a human personally said, did, experienced, or endorsed what the advertisement portrays.
• Materially alters a product demonstration or other evidence supporting an advertising claim.
• Would, if not disclosed, leave consumers with a materially different understanding of the advertisement.
• Triggers a state, platform, or foreign disclosure requirement regardless of the deception analysis.

These questions reflect a distinction between:

• AI assistance. Technology helps a human create content while the human remains the actual source of the relevant experience, opinion, or communication.
• AI authenticity substitution. Technology creates an apparent experience, opinion, speaker, or endorsement that does not actually exist.

The distinction is not absolute. A modest editing function can become significant depending on what it changes. However, the distinction focuses attention on the element of influencer marketing that AI most directly affects: authenticity.

Managing Risk

Internal AI Policies

Brands and agencies should consider adopting AI policies that:

• Identify approved AI tools and permitted uses, including uses by influencers and agency personnel.
• Classify uses by risk, distinguishing AI assistance from replicas and synthetic personas.
• Prohibit prompts that reference a real person by name or seek to imitate an identifiable person’s appearance or voice unless the person has consented in writing.
• Require human review of AI-generated consumer-facing content for factual accuracy, unsupported claims, hallucinations, and misleading implications, and compare AI output against a library of approved, substantiated claims.
• Require heightened review of AI edits to product demonstrations and influencer appearance, focusing on whether an edit changes what consumers are likely to believe about the product or its results.
• Add a question to creative review asking whether AI was used to create, modify, translate, or personalize the content, and how.
• Track state, platform, and foreign disclosure requirements by campaign and jurisdiction.
• Retain prompts, outputs, and approval records, which may be needed to substantiate claims or respond to a challenge.

Influencer Briefs and Agreements

Influencer briefs and agreements should:

• Address AI use expressly, including whether influencers may use GenAI to expand or rephrase approved talking points, and require review of any new claims the AI introduces. A practical rule is that AI may change the tone of approved claims but not their substance.
• Prohibit modifying content after approval, including through AI tools, without resubmitting it for review.
• Require influencers to confirm that first-person statements reflect their actual experience with the product.
• Address digital replica rights expressly and separately from general name, image, and likeness grants, including whether AI training, voice cloning, and synthetic content are permitted, what may be generated and for which campaigns, statements the replica may never make, approval rights, duration, post-termination use, retention or deletion of models and training data, and compensation. Provisions should include a reasonably specific description of intended uses, and brands should consider confirming that the influencer was represented by counsel, consistent with applicable state law (see, for example, Cal. Lab. Code § 927; N.Y. Gen. Oblig. Law § 5-302; 815 ILCS 550/1 et seq.).
• Allocate responsibility for platform labeling and state law disclosures.

Agency Agreements

Agencies may be liable as intermediaries for their roles in creating or disseminating deceptive endorsements (16 C.F.R. § 255.1(f)). Agency agreements should require the agency to disclose its use of AI tools for the brand’s work, route all AI-generated or AI-modified content through the brand’s legal review, comply with the brand’s AI policy, and allocate liability for AI-generated claims and unauthorized likeness or voice use through representations, warranties, and indemnities.

Replica and Synthetic Content Review

Before publishing replica or synthetic influencer content, brands and agencies should consider:

• Whether the influencer actually made the statements attributed to the influencer.
• Whether the influencer had the product experience portrayed, and still holds the views and uses the product as represented.
• Whether the synthetic content introduces claims or opinions beyond those the influencer actually expressed.
• Whether a reasonable consumer would understand the content to be a genuine recording or the persona to be a real person.
• Whether a synthetic persona resembles an identifiable real person.
• Where any AI disclosure is placed, and whether it will travel with the post when it is shared or recommended.
• Whether first-person statements by a virtual influencer, such as “I use this every morning” or “this worked for me,” describe experiences that do not exist.

Monitoring and Response

The Endorsement Guides expect advertisers to monitor their endorsers’ compliance and to remedy problems when they are discovered (16 C.F.R. § 255.1(d)). Monitoring should extend to content after it is approved, because AI tools make post-approval edits easy. If problematic AI-generated content is published, brands should remove it while preserving the post, prompts, outputs, and related communications, determine how the content was created and approved, assess whether corrective statements are needed, review agency and influencer contract remedies, and update their processes to prevent recurrence.

• Add a question to creative review asking whether AI was used to create, modify, translate, or personalize the content, and how.
• Track state, platform, and foreign disclosure requirements by campaign and jurisdiction.
• Retain prompts, outputs, and approval records, which may be needed to substantiate claims or respond to a challenge.

Conclusion

In the first two articles in this series, “AI, Influencers, and Liability: What Marketing Agencies Need to Know Before Deploying AI Tools” and “GenAI in Advertising and Influencer Marketing: The Legal Framework (2026 Update),” we explained that the key legal question for brands and agencies is not simply whether AI is being used in influencer marketing, but how it is being used. The three categories examined in this article bear that out. Each raises a different question: whether a real influencer remains the source of the experience and opinion, whether AI is putting words and conduct in a real person’s mouth, or whether a fictional persona is presenting a first-person experience that never happened. The common thread is the line between AI assistance, where a human remains the source of the experience or opinion, and AI authenticity substitution, where technology creates an experience, opinion, speaker, or endorsement that does not exist.

Federal law still does not require disclosure solely because AI was used, but the FTC applies existing deception, substantiation, and endorsement principles fully to AI-generated content. Endorsements must reflect genuine opinions and experiences, claims must be substantiated whether or not AI drafted them, and the Consumer Reviews and Testimonials Rule gives the FTC a civil penalty remedy for fabricated testimonials, including those delivered through a synthetic persona. At the same time, New York now requires disclosure of synthetic performers in advertising, states continue to expand digital replica protections and limit vague replica provisions in contracts, platforms require labels for realistic AI content, and the EU AI Act’s deepfake disclosure requirements now apply.

Brands and agencies should classify each use of AI by risk and adopt policies to match. Influencer agreements should address digital replica rights expressly and separately from general name, image, and likeness grants. A replica should not attribute statements or experiences to a real influencer that the influencer did not actually make or have, and a synthetic persona should not make a first-person claim that a human who never used the product could not truthfully make. Post-approval edits should be controlled, records of prompts and outputs retained, and state, platform, and EU requirements tracked campaign by campaign.

Those that evaluate AI based on what it does and how consumers will understand the result, rather than simply whether it was used, will be better positioned to use GenAI effectively while minimizing legal and reputational risk.

For further information and assistance, contact Lindsay Gehman or Saachi Gorinstein.

To view a PDF version of this article, please click here.

By Scott Hall and Meaghan Henderson

States continued to adopt more targeted privacy requirements in 2026, including new restrictions on precise geolocation and sensitive personal data and expanded requirements for data brokers. These laws are important even for businesses that have already determined whether they are subject to each state’s comprehensive privacy law because some requirements apply based on the type of data or activity involved rather than traditional consumer-volume or revenue thresholds.

Recent developments in Vermont, Maryland, Virginia, and New Jersey illustrate this broader shift toward regulating specific categories of data and data-processing activities.

Vermont Expands Data Broker Requirements

Vermont expanded its data broker requirements in 2026, adding new disclosure and financial requirements for registered data brokers.

The amendments expand required registration disclosures to include certain sensitive-data practices and disclosures to government bodies, law enforcement, foreign actors, and generative AI developers.

These requirements may be particularly relevant for data brokers that maintain detailed consumer profiles or provide personal information to multiple types of recipients. The specific inclusion of disclosures to generative AI developers also illustrates the increasing overlap between state privacy and AI regulation.

Vermont’s amendments also require registered data brokers to maintain a $20,000 bond for liabilities arising under the law.

Businesses that buy, sell, or license personal information should determine whether they qualify as a data broker under Vermont law and, if so, confirm that their registration disclosures and financial requirements reflect the amendments.

Maryland Expands Its Definition of Precise Geolocation

Maryland expanded the definition of precise geolocation information under its privacy law.

The definition now includes information identifying the location of a consumer, mobile device, or vehicle within a radius of 1,750 feet.

Businesses that collect location information should review whether data they previously treated as general location information may now fall within Maryland’s definition of precise geolocation information and become subject to additional restrictions.

Virginia Prohibits the Sale of Precise Geolocation Information

Virginia has adopted a more direct restriction on precise geolocation data by prohibiting businesses from selling or offering to sell precise geolocation information.

Businesses that collect location information should therefore understand not only whether they collect precise geolocation data, but also how that information moves through advertising, analytics, vendor, and other third-party relationships.

The Maryland and Virginia developments demonstrate the growing importance of treating location information as a distinct category when conducting data inventories and reviewing third-party data flows.

New Jersey Prohibits Sale of Sensitive Personal Data

New Jersey has taken an even broader approach to sensitive information by prohibiting the sale of sensitive personal data.

Significantly, the prohibition applies regardless of the number of New Jersey consumers whose information a company processes. A business may therefore fall outside the applicability thresholds of New Jersey’s comprehensive privacy law and still be subject to the sensitive-data sale prohibition.

As we discussed in our separate article on New Jersey’s new data broker law, the restriction applies broadly to sensitive information, including certain health, financial, biometric, precise geolocation, and other sensitive personal data.

What Businesses Should Do Now

These developments reinforce the need to look beyond the applicability thresholds of state comprehensive privacy laws. Businesses should:

• Map sensitive and location data. Identify sensitive personal data, precise geolocation information, biometric data, and other specially regulated information and understand how that information is used, sold, licensed, and shared.
• Review data sales and third-party disclosures. Determine whether advertising, analytics, licensing, or other data-sharing arrangements involve information subject to state-specific restrictions.
• Review data broker requirements separately. Businesses that buy, sell, or license personal information should evaluate data broker laws on a state-by-state basis rather than assuming their status under one state’s law determines their obligations elsewhere.
• Do not rely solely on comprehensive-law thresholds. Identify requirements that apply based on the type of data or processing activity even where the business does not otherwise meet a state’s general privacy-law thresholds.

The growing number of targeted state requirements means that a traditional state-by-state applicability analysis may not capture all of a business’s privacy obligations. Businesses should increasingly consider both where they operate and the specific types of data and processing activities involved.

Fall Privacy Webinar – November 17, 2026

Join Scott Hall and members of the Coblentz Data Privacy Team on Tuesday, November 17, 2026 for a timely discussion of the latest developments shaping the privacy, cybersecurity, and AI landscape and what they mean for your business. Learn practical steps to manage risk and gain insight into emerging state and federal AI developments; new and amended consumer privacy laws; enforcement trends involving children’s and sensitive data; and key privacy litigation and regulatory developments. RSVP for the webinar here.

If your company needs assistance with any privacy issues, Coblentz Data Privacy & Cybersecurity attorneys can help. Please contact Scott Hall at shall@coblentzlaw.com for further information or assistance.

By Scott Hall and Meaghan Henderson 

Connecticut amended its comprehensive privacy law in 2026, lowering the Connecticut Data Privacy Act’s (CTDPA) applicability thresholds and expanding requirements for sensitive data, artificial intelligence, profiling, automated decision-making, facial recognition, precise geolocation, and consumer rights, setting the state near the forefront of U.S. privacy laws.

Businesses that previously determined the CTDPA did not apply, as well as businesses that have not recently revisited their Connecticut privacy programs, should reassess their obligations under the new, more stringent standards.

More Businesses Will Be Subject to the CTDPA

The amendments lower the CTDPA’s general applicability threshold from 100,000 to 35,000 consumers.

They also replace the prior alternative threshold, which applied to businesses that processed the personal data of at least 25,000 consumers and derived more than 25% of gross revenue from the sale of personal data, with broader triggers tied to offering personal data for sale and processing sensitive personal data.

Sensitive Data and Consumer Protections Expand

Connecticut expands the types of information subject to heightened protections and the rights available to consumers. The amendments expand the definition of sensitive personal data and consumer access rights to include certain inferences generated from personal data and raise the age threshold for certain processing of minors’ data from 16 to 18 years old.

Under these amendments, businesses may no longer sell sensitive personal data without consent and are prohibited from selling the personal data of minors under 18, regardless of consent.

The amendments also narrow the definition of publicly available information and expand deletion rights for certain information obtained from public sources. Among other changes, information created by combining personal data with publicly available information, certain genetic data, and information posted where a consumer has a reasonable expectation of privacy may no longer qualify as publicly available. Consumers may also request deletion of publicly available information that is combined to create a publicly accessible consumer profile or made available for sale, as well as inferences generated from that information.

New Requirements for AI, Profiling, and Automated Decision-Making

Like California, Connecticut continues to expand its regulation of AI, profiling, and automated decision-making.

The amendments add a new disclosure requirement relating to artificial intelligence and expand consumer rights involving profiling and automated decision-making. These changes are relevant to businesses using AI and automated systems that generate inferences or other information about individuals or use those outputs in decisions affecting individuals.

Facial Recognition and Precise Geolocation Face Additional Restrictions

Separate Connecticut amendments effective October 1, 2026 add requirements relating to facial recognition technology and precise geolocation information.

The amendments restrict the sale of precise geolocation data and add protections addressing facial recognition technology. Businesses that collect or use this information may therefore have additional obligations depending on how the information is collected, used, disclosed, or sold and should pay particular attention to any new obligations that may arise.

Assessment Requirements Expand

Connecticut’s amended assessment requirements apply to processing activities created on or after August 1, 2026.

The amendments require assessments for certain processing activities, including activities involving sensitive data, profiling, automated decision-making, or other higher-risk processing. Businesses already conducting privacy, data protection, or AI assessments under other state laws may be able to leverage those processes to satisfy Connecticut’s requirements.

What Businesses Should Do Now

Businesses that operate in Connecticut or process Connecticut residents’ personal data should:

• Recheck applicability. Determine whether the lower thresholds or new applicability triggers bring the business within the scope of the CTDPA.
• Review sensitive and higher-risk data. Identify sensitive data, precise geolocation information, facial recognition information, and inferences collected or generated by the business.
• Review AI, profiling, and automated decision-making. Identify relevant activities and determine whether they trigger additional disclosure, consumer-rights, or assessment requirements.
• Review notices and consumer-rights procedures. Confirm that existing notices and procedures address the amended requirements, including changes relating to inferences and other newly regulated information.
• Review assessment requirements. Identify new or planned processing activities that require assessments and determine whether existing assessments can be leveraged.

Fall Privacy Webinar – November 17, 2026

Join Scott Hall and members of the Coblentz Data Privacy Team on Tuesday, November 17, 2026 for a timely discussion of the latest developments shaping the privacy, cybersecurity, and AI landscape and what they mean for your business. Learn practical steps to manage risk and gain insight into emerging state and federal AI developments; new and amended consumer privacy laws; enforcement trends involving children’s and sensitive data; and key privacy litigation and regulatory developments. RSVP for the webinar here.

If your company needs assistance with any privacy issues, Coblentz Data Privacy & Cybersecurity attorneys can help. Please contact Scott Hall at shall@coblentzlaw.com for further information or assistance.

By Scott Hall and Meaghan Henderson

On September 2, 2026, Delaware amended its comprehensive privacy law, the Delaware Personal Information Privacy Act (DPDPA), by signing House Bill 380 (HB 380) into law. HB 380, which goes into effect on January 1, 2027, marks the most significant revision to the DPDPA since its enactment in 2023.

HB 380 lowers the law’s applicability thresholds, narrows certain exemptions, creates new obligations for businesses that receive personal information from covered companies, and expands requirements governing sensitive information, profiling, and automated decision-making.

Businesses that previously determined the DPDPA did not apply, as well as businesses with existing Delaware privacy programs, should review their practices to determine whether the amended law changes their obligations.

More Businesses Will Be Subject to the DPDPA

HB 380 substantially lowers the DPDPA’s applicability thresholds:

• The general threshold decreases from 35,000 to 10,000 consumers annually.
• The threshold for businesses deriving more than 20% of gross revenue from the sale of personal information decreases from 10,000 to 5,000 consumers.
• Certain third parties that receive personal information from a controller may now be subject to the DPDPA without a minimum consumer threshold.

Financial Institution and Other Exemptions Are Narrowed

HB 380 narrows certain exemptions, including:

• The GLBA financial-institution exemption is narrowed. HB 380 replaces the broad entity-level exemption with more targeted exemptions for specified financial entities and activities, bringing Delaware closer to California’s approach. Information subject to the GLBA remains exempt, but the exemption no longer automatically extends to all affiliates and subsidiaries.
• The employee-information exemption is narrowed. Personal information used in connection with certain profiling and reports in the employment context will now be subject to the DPDPA.

Businesses that previously relied on an entity-level exemption should confirm that the exemption continues to apply after January 1, 2027.

Expanded Protections for Sensitive Information

HB 380 expands the definition of sensitive data to include the following categories, bringing it closer to the CCPA’s definition of sensitive personal information:

• Inferences used to reveal or identify sensitive characteristics, including health conditions, religious beliefs, racial or ethnic origin, sexual orientation, or immigration status;
• Neural information;
• Certain financial account credentials; and
• Certain government-issued identification numbers.

Controllers must obtain consent before processing sensitive information, and the processing must be reasonably necessary and proportionate to the purposes disclosed to the consumer.

Sales of sensitive data are also subject to additional restrictions. A permitted sale must be strictly necessary to provide or maintain a product or service requested by the consumer, and the controller must provide clear and conspicuous notice, obtain express consent, and maintain a record of that consent for five years.

New Vendor Due Diligence and Contracting Requirements

HB 380 adds requirements for businesses that disclose personal information to third parties:

• Controllers must enter into contracts with covered third parties that limit their use of personal information and include specified privacy protections, including audit and remediation rights.
• Controllers must conduct reasonable due diligence, including questionnaires and, where appropriate, additional measures based on the sensitivity of the data and the nature of the processing.
• Third parties must cooperate with those diligence efforts and provide information needed for the controller’s assessment.
• A third party without a required contract may not further process the personal data it receives and must comply with applicable contractual and diligence requirements.

Businesses should review vendor-management and contracting processes for relationships involving disclosures of personal information to third parties.

Profiling and Automated Decision-Making

HB 380 expands requirements for profiling and automated decision-making:

• New requirements apply to certain “reports,” including recommendations, summaries, automated decisions, and other communications based on personal information or profiling.
• Where a report is provided to a third party for use in a decision producing legal or similarly significant effects, the controller may need to ensure that the third party provides adverse-action notice, identifies the personal data relied upon, provides information about consumer rights, and offers human review where technically feasible.
• The right to opt out of profiling used in connection with automated decisions producing legal or similarly significant effects is expanded by removing the current “solely” automated-processing limitation.

These provisions may be relevant to businesses using AI or other automated tools in employment, financial services, insurance, housing, and other contexts involving significant decisions about individuals.

Assessment Requirements Also Expand

HB 380:

• Lowers the threshold for information protection assessments from 100,000 to 50,000 consumers; and
• Requires a separate impact assessment for profiling used in furtherance of automated decisions producing legal or similarly significant effects.

The profiling assessment must address, among other things, intended use, foreseeable harms and mitigation, data inputs and outputs, performance and transparency, post-deployment monitoring, and safeguards.

Businesses already conducting privacy or AI impact assessments under other state laws should determine whether those processes can be adapted to address Delaware’s requirements.

Other Changes to Consumer Rights and Privacy Practices

HB 380 makes several additional changes, including:

• Consumers generally will be entitled to identify the specific third parties that received their personal information, rather than only categories of third parties.
• Controllers must stop processing information subject to withdrawn consent within 15 days.
• Controllers must obtain consent from minors ages 13 to 18 before processing their personal information for profiling.
• Privacy notices must be reasonably particular to the product or service offered to the consumer, which may require businesses with materially different products or services to reconsider using a single general notice.

What Businesses Should Do Before January 1, 2027

Businesses operating in Delaware or processing Delaware residents’ personal information can use the remainder of 2026 to:

• Recheck applicability. Reevaluate the lower consumer thresholds, new third-party provisions, and existing exemptions.
• Identify sensitive information and inferences. Determine whether systems generate newly covered sensitive information and whether consent mechanisms remain sufficient.
• Review third-party relationships. Identify relevant disclosures and update contracts and diligence procedures.
• Inventory profiling and automated decision-making. Identify systems that may trigger the new requirements.
• Review assessment requirements. Determine whether existing or planned processing requires an information protection or profiling impact assessment and whether assessments prepared for other states can be leveraged.
• Update privacy notices and consumer-rights procedures. Confirm that notices, opt-out mechanisms, consent processes, and rights-request procedures address the amendments.

Fall Privacy Webinar – November 17, 2026

Join Scott Hall and members of the Coblentz Data Privacy Team on Tuesday, November 17, 2026 for a timely discussion of the latest developments shaping the privacy, cybersecurity, and AI landscape and what they mean for your business. Learn practical steps to manage risk and gain insight into emerging state and federal AI developments; new and amended consumer privacy laws; enforcement trends involving children’s and sensitive data; and key privacy litigation and regulatory developments. RSVP for the webinar here.

If your company needs assistance with any privacy issues, Coblentz Data Privacy & Cybersecurity attorneys can help. Please contact Scott Hall at shall@coblentzlaw.com for further information or assistance.

By Scott Hall and Meaghan Henderson

Less than a decade after California enacted the California Consumer Privacy Act (CCPA) in 2018, privacy legislation has spread to nearly half of the United States. With Alabama, Louisiana, Oklahoma, and Vermont joining the list this year, 24 states now have their own comprehensive consumer privacy laws, with at least three more regulating privacy through more targeted legislation.

With privacy a clear legislative priority across the country, businesses should expect privacy compliance to remain an increasingly important operational and legal issue. But the growing number of laws is not the only challenge.

Although the four new laws generally follow the framework established by existing state privacy laws, they differ in important ways. Businesses with multistate privacy programs should consider whether those differences require updates to their existing notices, consumer-rights procedures, data inventories, and risk-assessment processes.

Oklahoma Consumer Data Privacy Act, effective January 1, 2027

Oklahoma’s law applies to businesses that operate in Oklahoma or target its residents and either:

• control or process the personal data of at least 100,000 consumers; or
• control or process the personal data of at least 25,000 consumers and derive more than 50% of gross revenue from the sale of personal data.

“Consumer” excludes individuals acting in a commercial or employment context, and “sale” is limited to exchanges for monetary consideration only. The law also exempts financial institutions and data subject to the GLBA, HIPAA-covered entities, and business associates, nonprofits, and higher education institutions.

Consumers receive familiar rights to access, correct, and delete personal data and opt out of targeted advertising, sales, and certain profiling. Businesses must obtain consent before processing sensitive data and conduct data protection assessments for certain higher-risk processing. Comparable assessments prepared for other state laws may be used.

There are, however, several differences businesses should note. Oklahoma requires authentication for consumer rights requests, including opt-out requests, and does not require businesses to recognize universal opt-out mechanisms. Its definition of personal data also expressly includes pseudonymous data when it is used with additional information that reasonably links the data to an identified or identifiable individual. And while the law generally excludes photographs and audio or video recordings from its definition of biometric data, that exclusion does not apply when data generated from those materials is used to identify a specific individual.

The Oklahoma Attorney General has exclusive enforcement authority. Businesses receive 30 days to cure alleged violations, and statutory damages may reach $7,500 per violation.

Alabama Personal Data Protection Act, effective May 1, 2027

Alabama’s law applies to Alabama businesses, or businesses targeting Alabama residents, that:

• control or process the personal data of more than 25,000 consumers, excluding data processed solely for payment transactions; or
• derive more than 25% of gross revenue from the sale of personal data.

The law also exempts businesses with fewer than 500 employees that do not sell personal data, among other entity- and data-level exemptions.

The law provides familiar consumer rights, including access, correction, deletion, and copies of their personal data, as well as opt-outs from targeted advertising, sales of personal data, and certain profiling. Businesses must obtain consent to process sensitive data. Alabama also provides additional protections for minors: businesses generally must obtain consent before using the personal data of consumers ages 13 through 15 for targeted advertising or selling that data.

However, Alabama differs from the other new laws in several respects. Its definition of a “sale” is broader than monetary consideration, but it generally does not treat certain disclosures or transfers for analytics or marketing services as sales of personal data. Alabama’s law also does not require formal data protection assessments, does not require businesses to recognize universal opt-out signals, and does not provide a consumer right to appeal a decision on a rights request.

Enforcement is limited to the Alabama Attorney General, with a 45-day cure period that does not sunset. 

Louisiana Data Privacy Act, effective January 1, 2027

Louisiana’s law applies to businesses that operate in Louisiana or target its residents and meet any one of the following thresholds:

• annual gross revenue exceeding $25 million;
• annually processing, buying, receiving, selling, or commercially sharing the personal information of at least 75,000 Louisiana consumers, households, or devices; or
• deriving at least 50% of annual revenue from the sale of consumer personal information.

Because the revenue threshold applies independently of the amount of personal information processed, the law may reach some businesses, including certain B2B companies, that would not otherwise expect to fall within a state comprehensive privacy law.

As with Oklahoma and Alabama, the law provides familiar consumer rights, including access, correction, deletion, portability, and opt-outs from targeted advertising, sales of personal data, and certain profiling. Although Louisiana does not provide consumers with the right to revoke consent, businesses must obtain consent before processing sensitive data and must conduct data protection impact assessments for certain higher-risk processing. Similar to Oklahoma, Louisiana also permits businesses to rely on comparable assessments prepared to comply with other state privacy laws.

There are several other Louisiana-specific requirements to be aware of. The law requires authentication for consumer rights requests, including opt-outs, and requires recognition of universal opt-out mechanisms. The privacy policy must also include specific disclosures if a business sells sensitive personal data or biometric personal data. Unlike Texas, Louisiana does not expressly include certain pseudonymous data within the definition of personal data.

The law is enforceable only by the Louisiana Attorney General and provides a 30-day cure period, although that cure period sunsets on July 31, 2027. 

Vermont Data Privacy and Online Surveillance Act, effective January 1, 2028

Vermont’s law is broader in several respects than the other state privacy laws enacted in 2026. It applies to Vermont businesses that:

• control or process the personal data of at least 35,000 consumers;
• control or process the sensitive data of at least 3,000 consumers; or
• offer for sale the personal data of at least 3,000 consumers.

Vermont also provides narrower exemptions than many other states, including no blanket nonprofit exemption and a narrower financial-institution exemption.

Consumers receive the familiar rights to access, correct, delete, and obtain copies of personal data and to opt out of targeted advertising, sales, and certain profiling. Vermont goes further, however, by providing additional rights relating to profiling and automated decisions, including rights in certain circumstances to question a profiling result, obtain information about the reasons for the decision, and review the personal data used. Businesses must also conduct data protection assessments for certain higher-risk processing and separate impact assessments for certain profiling used to make decisions with legal or similarly significant effects.

The law also contains several requirements that businesses should flag for their existing compliance programs. Vermont uses a broader definition of biometric data that does not require the data to be processed for purposes of uniquely identifying an individual. Its definition of publicly available information is also narrower and generally excludes information aggregated into a consumer profile and inferences derived from that profile. In addition, privacy policies must disclose whether the business collects, uses, or sells personal data to train large language models.

Vermont provides additional protections for minors and consumer health data. Businesses generally may not sell personal data or use it for targeted advertising when they have actual knowledge, and willfully disregard, that a consumer is between 13 and 17 years old. The law’s consumer health data requirements are particularly broad because they apply without a minimum processing threshold and restrict, among other things, access to consumer health data, certain geofencing around health care facilities, and sales of consumer health data without consent.

The Vermont Attorney General has exclusive enforcement authority. Businesses receive 60 days to cure violations during the first 18 months after the law takes effect.

Common Requirements and Important Differences

The four laws generally follow the now-familiar state privacy framework: covered businesses must provide privacy notices, honor consumer rights, and address sensitive data, targeted advertising, sales of personal information, and certain profiling.

The laws differ in important ways that may require updates to existing multistate privacy programs.

Applicability thresholds vary substantially, from Oklahoma’s 100,000-consumer threshold and Alabama’s 25,000-consumer threshold to Louisiana’s $25 million revenue threshold and Vermont’s separate thresholds for personal data, sensitive data, and data offered for sale. The states also take different approaches to nonprofit and small-business exemptions, universal opt-out mechanisms, authentication of consumer rights requests, and the treatment of pseudonymous and biometric data.

The laws also differ in their assessment requirements. Louisiana, Oklahoma, and Vermont require data protection assessments for specified higher-risk processing, and Vermont separately requires impact assessments for certain profiling activities. Alabama does not require formal data protection assessments. Businesses should also consider whether existing assessments can be used to satisfy the requirements of multiple states.

Finally, businesses should not assume that a single privacy program will address all four laws without modification. Vermont in particular adds requirements relating to profiling, minors, consumer health data, biometric data, and disclosures concerning the use of personal data to train large language models. These differences reinforce the need to review existing privacy notices, consumer-rights procedures, data inventories, assessment processes, and other compliance controls as new state laws take effect.

The continued expansion of state privacy laws makes a consistent multistate framework increasingly valuable, but meaningful differences among states still require businesses to evaluate their practices state by state.

Fall Privacy Webinar – November 17, 2026

Join Scott Hall and members of the Coblentz Data Privacy Team on Tuesday, November 17, 2026 for a timely discussion of the latest developments shaping the privacy, cybersecurity, and AI landscape and what they mean for your business. Learn practical steps to manage risk and gain insight into emerging state and federal AI developments; new and amended consumer privacy laws; enforcement trends involving children’s and sensitive data; and key privacy litigation and regulatory developments. RSVP for the webinar here.

If your company needs assistance with any privacy issues, Coblentz Data Privacy & Cybersecurity attorneys can help. Please contact Scott Hall at shall@coblentzlaw.com for further information or assistance.

By Scott Hall and Katie Berens

The U.S. Court of Appeals for the Ninth Circuit recently issued a novel ruling that an agentic AI tool did not trigger liability under the Computer Fraud and Abuse Act (CFAA) and the California Comprehensive Computer Data Access and Fraud Act (CDAFA)—statutes that often appear as companion claims in privacy cases involving unauthorized access to computers, devices, or data.

On August 4, 2026, in Amazon.com Services, LLC v. Perplexity AI, Inc., No. 26-1444, the Ninth Circuit vacated the trial court’s award of a preliminary injunction to Amazon, holding that Amazon was unlikely to succeed on the merits of its CFAA and CDAFA claims because it could not show Perplexity’s agentic AI tool “accessed” Amazon computers, a requirement of both statutes.

While the Court expressly limited its narrow ruling, this first-of-its-kind decision offers a glimpse into how courts may approach agentic AI issues in similar contexts, and raises broader questions about when agentic AI may trigger liability for statutory or other privacy-related violations.

Background

This case involved Perplexity’s AI-enabled web browser, Comet, which it developed and publicly released. Comet runs locally on a user’s machine and includes an AI agent (which the Court refers to in its opinion as “the Assistant”) that performs tasks at the user’s direction, including online shopping.[1]

Amazon allegedly told Perplexity, prior to Comet’s release, that its AI products would not be allowed to access the Amazon.com store.[2] After Perplexity released Comet, Amazon detected that the Assistant was shopping its store on behalf of Comet users, and Amazon again informed Perplexity that the Assistant was not authorized to use the store or make purchases for its customers.[3]

Amazon then filed suit against Perplexity for violations of the CFAA and the CDAFA and moved for a preliminary injunction to prohibit the use of Perplexity’s agentic web browser AI tool on Amazon’s website.[4] Amazon claimed that Perplexity violated the CFAA, a federal anti-hacking statute, and its California state-law analogue, the CDAFA.[5] Both are criminal statutes that provide a private right of action for civil claims. The CFAA prohibits intentionally accessing and obtaining information from computers without authorization, and makes relief available to plaintiffs who have suffered damage or loss of at least $5,000 per year as a result of a violation.[6] The CDAFA similarly prohibits knowingly accessing without permission any computer system or network, but it does not impose a minimum loss threshold.[7]

Amazon claimed that by way of the Assistant, Perplexity’s servers had accessed Amazon’s servers without Amazon’s authorization.[8] Perplexity countered that its own computers never gained “access” to any Amazon computer or server, because when the user deploys the Assistant, any Amazon data is first transmitted to the user’s computer—and any “intent” would belong to the user, not Perplexity, as the Assistant is computer software.[9]

The Court Rejects Amazon’s Argument

The Court ruled in favor of Perplexity, vacating the preliminary injunction and concluding that Amazon was unlikely to succeed on the merits because it failed to show that Perplexity “accessed” Amazon’s servers as required under the CFAA and the CDAFA.[10] The Court determined that the evidence did not support Amazon’s argument that Perplexity itself directly communicated with Amazon’s servers, as it is the user who controls the Assistant to carry out shopping tasks on Amazon.com.[11] It further explained that the plain language of the CFAA does not allow for “access” by the Assistant, as “access” only applies to a person—and “[h]owever advanced the Assistant currently is, it is a tool, not a person for statutory purposes.”[12]

The Court reiterated that agentic AI is an emerging technology, and that this decision does not “establish a new legal regime governing agentic AI”—nor does it address the potential for liability in other contexts such as tort claims.[13] It also acknowledged that “[t]he legal understanding of agentic AI will doubtless change as AI technology grows increasingly sophisticated.”[14] However, this case represents an early marker in what is sure to become a hotly contested issue going forward.

Key Takeaways

As the Court noted, this ruling is limited and does not insulate agentic AI developers or deployers from potential liability under these statutes or other laws, nor from claims based on violations of a website’s terms of service.

It does, however, provide a novel illustration of a court conducting a piecemeal analysis of how an agentic AI tool operates to determine liability, including an examination of who controls—and may ultimately be responsible for—the actions that the agent undertakes. As this case shows, each outcome may depend on highly factual questions, including how the tool is built and whether the agent is acting only on behalf of the user.

As the law evolves to keep pace with the rapid advancement of agentic AI, companies should stay informed about key judicial decisions and legislative activity and their potential implications for the AI tools that they develop and deploy.

Fall Privacy Webinar – November 17, 2026

Join Scott Hall and members of the Coblentz Data Privacy Team on Tuesday, November 17, 2026 for a timely discussion of the latest developments shaping the privacy, cybersecurity, and AI landscape and what they mean for your business. Learn practical steps to manage risk and gain insight into emerging state and federal AI developments; new and amended consumer privacy laws; enforcement trends involving children’s and sensitive data; and key privacy litigation and regulatory developments. RSVP for the webinar here.

If your company needs assistance with any privacy issues, Coblentz Data Privacy & Cybersecurity attorneys can help. Please contact Scott Hall at shall@coblentzlaw.com for further information or assistance.

[1] Amazon.com Services, LLC v. Perplexity AI, Inc., 184 F.4th 1083, 1087–88 (9th Cir. 2026).

[2] Id. at 1088.

[3] Id.

[4] Id.

[5] Id.; 18 U.S.C. § 1030; Cal. Penal Code § 502.

[6] 18 U.S.C. § 1030(a)(2), (g).

[7] Cal. Penal Code § 502. For more information on the recent rise in CDAFA claims in the privacy context, including with regard to website tracking technologies, see our prior article here.

[8] Id. at 1089–90.

[9] Id. at 1089.

[10] Id. at 1090–91.

[11] Id. at 1091.

[12] Id.

[13] Id. at 1092.

[14] Id.

By Scott Hall 

Businesses can no longer wait to take AI compliance seriously. While there is still no federal AI statute, more than 60 state laws now regulate various aspects of commercial AI, and there is already substantial litigation involving business use of AI tools, largely brought under privacy, wiretap, and biometric statutes that predate the technology by decades. The good news for businesses is that most of what these new AI regimes require is familiar: inventory, notice, consent, vendor terms, retention limits, and human review.

If your business has not evaluated its own AI compliance, both in terms of the tools used internally and any products, services, or tools deployed as part of its business operations, now is the time to do so.

1. Laws to Know

Federal

Recent federal policy has moved toward light-touch AI regulation and a single national framework. The current administration is challenging state AI laws and pressing Congress to preempt them, but that effort has not yet displaced any state laws. Taking a lesson from privacy law, businesses should not take a wait-and-see approach in reliance on a future federal AI regime or preemption that may never happen. For now, companies should make plans to comply with state AI and privacy laws and rules that apply to their products, users, and data.

Notably, the FTC can act under Section 5 of the FTC Act even when no AI-specific statute applies. Businesses can face enforcement for undisclosed or misleading AI practices, including sharing user data with outside AI developers in a way its privacy disclosures do not explain.

States–Four Primary Areas of Legislation

• Chatbots—the most active area of AI lawmaking. California, New York, Connecticut, Colorado, Washington, and several other states have adopted chatbot laws and rules. The details vary, but three obligations are common: (1) disclose when a user is interacting with AI, (2) maintain protocols to detect and respond to self-harm concerns, and (3) protect minors.Coverage turns on how a state defines a chatbot. Some rules reach ordinary customer-service bots, while others focus on companion systems that simulate relationships or remember prior interactions. These laws generally apply based on where your users are located, not where you are incorporated, so a nationwide chatbot should be built to the strictest applicable standard.
• Automated decisions about people. New AI laws and rules are focusing on significant decisions made by AI or automated means. If AI helps decide who gets hired, fired, admitted, housed, insured, or provided services, the company using it should be prepared to disclose the use, explain the decision process, and document testing and oversight. Where states are splitting is the threshold question: how much AI involvement triggers those duties.California’s CCPA rules reach technology that replaces or substantially replaces human decisionmaking, meaning the output is used without meaningful human involvement. A reviewer who understands the output, weighs other relevant information, and has authority to change the result keeps the use outside the rules. Rubber-stamping does not.Colorado’s proposed SB 26-189 sets a broader trigger: AI that materially influences a consequential decision in seven areas including employment, housing, lending, insurance, and healthcare. The proposed law would also require notice after an adverse outcome and let individuals correct the underlying data and seek human reconsideration. The statute does not define “materially influences,” and the Attorney General has asked for comment on competing standards.The practical consequence: keeping a human in the loop may take you outside some state laws but not others. Inventory where AI informs decisions about people, not only where it makes them on its own.

• Sectoral laws, content regulation, and AI-as-privacy. Specific review of AI laws based on industry sector is important, as states continue to regulate AI more restrictively in contexts such as healthcare, employment, and insurance, requiring express notice and consent and restricting uses in various contexts. States are also regulating AI content, including restricting deepfakes, digital replicas, and AI use in election contexts. And many states, including California, continue to regulate many aspects of AI as part of privacy law compliance.
• Developer and frontier-model laws. Finally, we continue to see state laws governing large-scale AI systems. California, New York, and Illinois are imposing safety frameworks, incident reporting, and independent review requirements on developers of powerful frontier AI systems. And while this may not apply to most businesses, everyone who deploys or uses large AI models should monitor these trends.

California-Specific Laws

The CPPA’s rules on automated decisionmaking technology (ADMT) and privacy risk assessments create the clearest near-term deadlines for AI compliance for many businesses:

• ADMT—January 1, 2027. If you use automated technology to make or materially influence significant decisions about California residents (employment, lending, housing, education, healthcare, or essential services), you need published pre-use notices, working opt-out and access workflows, and an explanation of the logic and effects. Employees, applicants, and contractors are included beyond commercial consumers.
• Risk assessments—underway since January 1, 2026. As of now, businesses subject to the CCPA must conduct risk assessments for high-risk processing activities and must submit attestations and summaries to the CPPA by April 1, 2028. High-risk processing includes ADMT uses.

California’s Legislature closed its 2026 session with a package of AI and privacy measures, including several laws aimed at regulating business/employer use of AI:

• SB 947—Bars sole reliance on automated systems for employee discipline or termination; requires human corroboration, post-use notice, and employee data access. Operative July 1, 2027.
• SB 951—Amends Cal/WARN to require additional disclosures when layoffs are caused substantially by AI or automation.
• AB 1883—Categorically prohibits employer use of AI tools to infer emotional states or collect neural data, with no wellness-program exception.
• AB 1331—Prohibits workplace surveillance tools in restrooms.
• Other measures include SB 923 (expanded deletion rights), AB 1609 (chatbot disclosures), and additional rules on data sharing and professional AI use.

The EU AI Act

If your business offers AI-enabled products or services to people in the EU, the EU AI Act can reach you without an EU entity. The EU AI Act applies based on where a system is placed on the market or where its output is used. Compliance obligations scale with risk, and the heaviest requirements fall on “high-risk” uses, which for most US companies means hiring and HR screening tools, credit decisions, and education or admissions systems.

2. Litigation and AI Tool Trends

AI notetakers are a current hot spot. If your company uses an AI notetaker, treat the tool as more than a transcription service. In August 2026, a California court allowed claims against Otter.ai to proceed after plaintiffs alleged that its speaker-tagging feature created voice profiles without the required consent. The same theory has been asserted against other tools, including one tool that took notes without visibly joining the meeting. Before deployment, confirm what data or inputs the tool/vendor records, how long it keeps the data, and whether it uses it for anything beyond providing the service. If so, a court may view it as a separate party and not merely your service provider, and thus subject to heightened notice and consent obligations.

Exposure is not limited to the AI tool/vendor. A company that deploys the tool may also be named, particularly if it helped decide how recordings or voice data were collected, used, or retained. Illinois’s BIPA allows private suits and statutory damages of $1,000 per negligent violation and $5,000 per intentional or reckless violation, or actual damages if greater. Recent class actions involving voiceprints from public recordings show how quickly these theories can spread. Speaker identification, voice authentication, and emotion detection are all risk areas.

Chatbots and website tracking. Claims under California’s CIPA involving website chat functions, session replay, and tracking pixels continue to increase and provide statutory damages of $5,000 per violation with no proof of harm required. California’s SB 690 would limit some pen-register and trap-and-trace claims involving websites and apps, but it does not eliminate the broader litigation risk under wiretapping/electronic communication legal theories.

Employment. Using a well-known AI hiring tool or vendor does not shield an employer from discrimination claims. AI screening tools used on an employer’s behalf can expose the employer to liability. Businesses using these tools should run bias audits through counsel where appropriate, and document the scope, methodology, and results.

AI-driven pricing. Dynamic pricing, which involves adjusting prices based on supply, demand, time of day, or competitor pricing, is a long-established business practice, and AI generally makes it faster and more efficient. But risk arises when a system uses personal consumer data, such as browsing behavior, purchase history, location, income proxies, or demographic information, to estimate an individual’s willingness to pay and set a price accordingly. That is personalized, or “surveillance,” pricing.

In August 2026, the FTC proposed an enforcement policy statement making these practices a key focus. Its position is that a business using consumer data to personalize prices should clearly disclose that the price is personalized, the basis for the personalization, and the types of data used; failing to do so may violate Section 5 as deceptive or unfair.

State consumer-protection statutes, privacy laws, and emerging state personalized-pricing legislation create additional risk, and models trained on historical data may produce prices that correlate with race, income, age, or other protected characteristics, creating civil-rights liability even without discriminatory intent. For business clients, the practical line is straightforward: dynamic pricing based on market conditions is generally permissible; personalizing prices from individual consumer data without disclosure is where the legal risk sits. If your business uses AI in pricing, review what data feeds the model and whether consumers know about it.

AI in Litigation and Discovery

Courts are treating AI prompts and outputs as ordinary business records rather than a protected category. For business teams, that means assuming employee chats, meeting transcripts, and AI tool logs may be requested in litigation.

• Prompts entered into public AI tools are generally not confidential. Asking a chatbot a legal question does not create an attorney-client communication, and using a consumer-grade tool undercuts any later claim that the exchange was privileged.
• Employee prompts can become evidence against the company. Courts have relied on chatbot conversations to establish what decision-makers were actually thinking, including where the stated business rationale differed from the exchange.
• Deleting AI logs creates a second problem. Once litigation is reasonably anticipated, AI records carry the same preservation obligations as email, and destroying them invites spoliation arguments on top of whatever the underlying dispute is.
• Work product protection may survive when counsel directs the use. Where an attorney structures the prompts or supervises the work, protection is more likely — but the law is unsettled, and the conservative assumption is that prompts, outputs, transcripts, and logs are discoverable unless counsel has directed otherwise.
• Update preservation infrastructure now. Litigation holds, custodian interviews, and discovery templates should expressly cover AI prompts and outputs, meeting transcripts and summaries, and related metadata — categories most existing templates do not name.
• Recording expands the record. Every AI-generated transcript or summary is a new discoverable document, and one that may not match the formal minutes of the same meeting.

Regulators are Active

State attorneys general are leading AI enforcement by using consumer-protection, civil-rights, and privacy laws that already exist. A coalition of AGs recently warned major AI companies about harmful outputs, and individual states have pursued chatbot operators and AI developers through investigations, injunctions, criminal referrals, and new legislation. Businesses should take note that regulators do not need to wait for AI-specific statutes before acting. They are adapting current laws to these new technologies.

3. Internal AI Use: Risks and Safeguards

Businesses should start from the assumption that AI is already in your workplace. Global survey data puts regular, intentional AI use at roughly three in five workers, and vendors are enabling AI features inside software you already license, sometimes without explicit notice. It is unlikely any business can still represent that they do not use AI, and broad AI tool bans are no longer practical.

Adopt a policy, but do not impose a ban. Prohibition on use of AI tools drives AI use underground, where it cannot be supervised. A workable policy identifies approved and prohibited tools and uses, states what may never be entered into an AI tool, addresses confidentiality and privacy, requires verification, assigns accountability for final work product, and explains consequences. Without a policy, every employee sets their own rules, with many using AI quietly and never disclosing it. Businesses should take a practical approach and reasonably regulate internal AI use. AI policy updates and employee trainings should be expected regularly as AI tools and the legal landscape continue to evolve rapidly.

Watch for AI tools that act, not just advise. Agentic AI systems can approve transactions, send communications, screen applicants, manage workflows, or make decisions with limited human involvement. These tools are arriving quickly, often embedded in software businesses already use, and the compliance stakes rise when the system acts rather than merely provides information. Inventory where AI is taking actions and ensure that appropriate human oversight and audit trails scale with the tool’s level of autonomy.

Use enterprise tools, not free or consumer accounts. Businesses should contract for closed AI systems that prohibit vendor retention, training on your data, and third-party disclosure. The gap between a company’s enterprise account and an employee’s personal free account now has real, legal consequences. Policies should make the distinction explicit and enforce it.

Make a deliberate decision about meeting assistants. We previously addressed these tools here, but the information bears mentioning again. Treat these tools as communications-capture tools rather than productivity :

• Designate meetings where they are never permitted—privileged advice, investigations, HR and accommodation discussions, board deliberations, transactions, and anything involving client-confidential, patient, or applicant information.
• Provide participant-facing notice and capture consent, keeping in mind all-party consent states and external participants who never agreed to your vendor’s terms.
• Disable speaker identification where a plain transcript will suffice. That setting is often the difference between a transcript and a biometric identifier.
• Confirm that administrators can set defaults centrally instead of leaving recording, sharing, and training settings to individual users.
• Reconcile AI summaries against formal minutes. Discrepancies get scrutinized, and verbatim transcripts preserve offhand remarks that curated minutes would omit.

Set retention and deletion rules. Don’t just let AI tool data sit in your systems. Set retention rules and deletion schedules for each category of data these tools generate, including the audio, transcripts and summaries, voice profiles, associated metadata, downloaded exports, and copies that linger in CRM or collaboration systems even after the original is deleted.

Maintain human oversight for important AI-generated output. Several new laws now require it, and many business contracts are now demanding it.

4. Three Priorities Before January 1, 2027

1. Build your AI inventory. You cannot govern what you have not mapped. Document every AI tool, system, and AI-enabled feature in use, as well as the data it touches, and the risk tier.
2. Map your exposure. Based on your AI inventory, identify applicable laws and legal requirements. California’s ADMT compliance, Colorado’s replacement automated-decision law, new state chatbot requirements, and many other laws are going into effect January 1, 2027, and many other laws are already in effect.
3. Name one accountable owner. When legal, HR, security, and the IT function each hold a piece of ownership, the seams between them are where problems go unnoticed. A single senior executive with a clear escalation path closes those seams.

The Bottom Line

AI can no longer be ignored. Whether or not your company has formally adopted it, AI is almost certainly already in use—in tools your employees chose, and in features your vendors switched on without asking. The starting point is not a policy. It is a set of questions every business should be able to answer:

• What AI are we using, including features enabled by default in software we already license?
• Have we disclosed that use to the people it affects—customers, employees, applicants, and meeting participants?
• What do our AI vendors do with our data, and can we turn it off?
• How long do we keep AI-generated content, and can we delete it when we need to?
• Who inside the business owns AI compliance and has authority to act?

Few companies can answer all five questions today. The ones that can will find most new obligations are variations on work they have already done—and the ones that cannot will likely discover the gaps through a regulator’s inquiry or a plaintiff’s discovery request. Ensuring your business can answer these questions now will put you in a good position to deal with the continuing evolution of AI laws and regulations.

Fall Privacy Webinar – November 17, 2026

Join Scott Hall and members of the Coblentz Data Privacy Team on Tuesday, November 17, 2026 for a timely discussion of the latest developments shaping the privacy, cybersecurity, and AI landscape and what they mean for your business. Learn practical steps to manage risk and gain insight into emerging state and federal AI developments; new and amended consumer privacy laws; enforcement trends involving children’s and sensitive data; and key privacy litigation and regulatory developments. RSVP for the webinar here.

If your company needs assistance with any privacy issues, Coblentz Data Privacy & Cybersecurity attorneys can help. Please contact Scott Hall at shall@coblentzlaw.com for further information or assistance.

 

 

 

 

Navigating the Evolving Legal Landscape of Data Privacy, Cybersecurity, and AI

By Scott Hall, Phillip Wiese, Meaghan Henderson, Katie Berens, Kat Gianelli, and Dillon Ostlund

Download a PDF version of this report here.

For businesses, the rules governing AI and personal information are being rewritten in real time, and the cost of falling behind has never been higher. AI touches nearly every aspect of daily life and has become near ubiquitous in businesses across the country. At the same time, consumers are becoming increasingly savvy about their individual privacy rights, demanding more information and stronger protections from the businesses to which they provide personal information. Reflecting the importance of these two issues, states are rapidly enacting, updating, and amending their privacy and AI laws. There are now 24 states with comprehensive consumer privacy laws, with piecemeal, topic- or industry-specific privacy and AI laws popping up across the country as well.

Children’s data and other sensitive data collection and use are particular areas of concern and focus. The federal government continues to weigh federal legislation, but has not, to date, passed any federal laws on either AI or privacy, leaving enforcement largely to the states and their patchwork of requirements.

Adding to the complexity, courts and regulators are applying decades-old statutes to new technologies, and regulators and private plaintiffs are more active than ever in examining and testing the bounds of privacy laws and consumer privacy rights. The result is a legal landscape more complex than ever before. What are businesses to do? How do businesses keep up with the ever-changing legal landscape?

Our 2026 Fall Privacy Report untangles the web of recent AI and privacy laws and provides concrete takeaways to help businesses navigate this landscape. View the full report here.

Fall Privacy Webinar – November 17, 2026

Join Scott Hall and members of the Coblentz Data Privacy Team on Tuesday, November 17, 2026 for a timely discussion of the latest developments shaping the privacy, cybersecurity, and AI landscape and what they mean for your business. Learn practical steps to manage risk and gain insight into emerging state and federal AI developments; new and amended consumer privacy laws; enforcement trends involving children’s and sensitive data; and key privacy litigation and regulatory developments. RSVP for the webinar here.

Jennifer Scharre co-authored “Proposition 40: California Billionaires in the Crosshairs” which was published in the California Trusts & Estates Quarterly. The article discusses Proposition 40, the “One-Time Wealth Tax for State-Funded Healthcare, Education, and Food Assistance Programs Initiative,” also known as the 2026 Billionaire Tax Act (the “Act”) and potential implications of the tax for taxpayers with the text of the Act as a guide. The full article is linked here.

By Lindsay Gehman and Saachi S. Gorinstein

Our June 2026 article “AI, Influencers, and Liability: What Marketing Agencies Need to Know Before Deploying AI Tools,” surveyed the legal risks organizations face when they use or develop GenAI, including risks under consumer fraud and unfair and deceptive acts and practices (UDAP) laws and under state rights of publicity.

Since then, the legal landscape for AI-generated marketing content has changed considerably. Federal enforcement priorities have shifted, a federal executive order has targeted state AI regulation, New York has adopted the first state law requiring disclosure of AI-generated performers in advertising, digital replica protections have expanded in several states, and the EU AI Act’s transparency obligations now apply.

This article updates and expands the earlier discussion as it applies to advertising and endorsements. It focuses on the rules most relevant to brands, marketing agencies, and influencers that use GenAI to create or modify marketing content, including more details on federal endorsement and review rules.

GenAI is used in influencer marketing in three principal ways:

• AI-assisted real influencers. A real influencer remains the speaker and source of the endorsement, while GenAI helps draft, edit, translate, or produce the content.
• AI replicas of real influencers. AI creates content depicting an identifiable real person saying or doing something the person did not actually record, such as through voice cloning, digital replicas, or synthetic video.
• Fully synthetic influencers. AI creates a fictional persona, sometimes with a realistic appearance, voice, and social media presence, that does not correspond to any real person.

This article sets out the legal framework that applies across all three categories.

A recurring question runs through that framework: when must consumers be told that AI was used? There is currently no general federal rule requiring disclosure solely because AI was used to create or edit advertising content. However, existing consumer protection law applies regardless of whether content is created by a human, an AI system, or both. Under federal law, the more relevant question is whether the use of AI changes what consumers are likely to believe about an advertisement, particularly its source, its authenticity, or the experience of the person making an endorsement. AI-specific disclosure obligations are also beginning to emerge in state law, platform policies, and the EU AI Act.

What Has Changed

The most significant developments include:

• FTC enforcement posture. In December 2025, the FTC reopened and set aside its 2024 consent order against AI writing-tool provider Rytr LLC, citing the July 2025 AI Action Plan. In July 2026, as directed by Executive Order 14365, the FTC proposed a policy statement on the application of Section 5 to AI systems. The FTC also finalized its order against Workado, LLC over unsupported AI-detection accuracy claims (see FTC Act Section 5).
• Federal action on state AI laws. Executive Order 14365, issued December 11, 2025, created a Department of Justice AI Litigation Task Force to challenge state AI laws deemed inconsistent with a national policy framework (see State AI Disclosure Laws).
• New York synthetic performer disclosure. Since June 9, 2026, advertisements that include AI-generated synthetic performers must disclose that fact under New York law (see State AI Disclosure Laws).
• Utah codification. Utah’s 2025 GenAI consumer protection amendments, including the rule that GenAI’s involvement is not a defense to a consumer protection violation, are codified in Chapter 77 of Title 13 of the Utah Code (see State UDAP Laws and State AI Disclosure Laws).
• California AI Transparency Act. California’s provenance disclosure law for large GenAI providers became operative on August 2, 2026 (see State AI Disclosure Laws).
• Expanded digital replica protections. In December 2025, New York broadened its protections for digital replicas of deceased performers. A federal court allowed New York right of publicity claims against an AI voice-cloning company to proceed. Additional states have enacted digital replica laws (see Rights of Publicity and Digital Replica Laws).
• NO FAKES Act. The federal digital replica bill was reintroduced as the NO FAKES Act of 2026, approved by the Senate Judiciary Committee, and placed on the Senate calendar (see Rights of Publicity and Digital Replica Laws).
• Impersonation Rule. The FTC has not finalized its proposal to extend the Impersonation Rule to individuals, has abandoned its proposed means-and-instrumentalities provision, and has opened a separate rulemaking inquiry about platform tools used in impersonation scams (see Impersonation Rule).
• EU AI Act transparency obligations. The EU AI Act’s Article 50 transparency obligations, including deepfake labeling, have applied since August 2, 2026 (see International Laws).

FTC Act Section 5

Section 5 of the FTC Act prohibits unfair or deceptive acts or practices in or affecting commerce (15 U.S.C. § 45(a)(1)). A representation, omission, or practice is deceptive if it is likely to mislead consumers acting reasonably under the circumstances and is material (FTC Policy Statement on Deception, appended to Cliffdale Assocs., Inc., 103 F.T.C. 110, 174 (1984)). Courts evaluate the net impression an advertisement creates, so a communication can be deceptive even if it contains truthful disclosures (FTC v. Cyberspace.com LLC, 453 F.3d 1196, 1200 (9th Cir. 2006); Kraft, Inc. v. FTC, 970 F.2d 311, 314 (7th Cir. 1992)).

Advertisers must also have a reasonable basis for objective claims before disseminating them. Claims that state or imply a specific level of support, such as “clinical research proves,” require at least that level of substantiation (FTC Policy Statement Regarding Advertising Substantiation, appended to Thompson Med. Co., 104 F.T.C. 648, 839 (1984), aff’d, 791 F.2d 189 (D.C. Cir. 1986); see also FTC, Health Products Compliance Guidance (Dec. 2022)). These obligations apply equally to claims drafted or suggested by GenAI. A GenAI tool may produce confident but unsupported statements, and the fact that a claim originated with AI does not reduce the advertiser’s substantiation obligation.

FTC Enforcement Involving AI

The FTC has repeatedly stated that existing law applies to AI-enabled conduct and has taken action against organizations that use AI, or claims about AI, to deceive consumers (see, for example, FTC: FTC Announces Crackdown on Deceptive AI Claims and Schemes (Sept. 25, 2024)). In April 2025, the FTC announced a proposed order requiring Workado, LLC to stop advertising the accuracy of its AI content detection products unless it has competent and reliable evidence to support those claims, and the FTC later finalized the order (FTC: FTC Order Requires Workado to Back Up Artificial Intelligence Detection Claims (Apr. 28, 2025)).

Under its current leadership, however, the FTC has framed AI enforcement as applying existing fraud and deception law rather than creating AI-specific obligations. The July 2025 AI Action Plan directed the FTC to review prior AI-related investigations, orders, and consent decrees that may unduly burden AI innovation (The White House, Winning the Race: America’s AI Action Plan (July 2025)). Consistent with that directive, in December 2025 the FTC reopened and set aside its 2024 consent order against AI writing-tool provider Rytr LLC, concluding that the complaint did not adequately allege a Section 5 violation and that the order unduly burdened AI innovation (FTC: FTC Reopens and Sets Aside Rytr Final Order in Response to the Trump Administration’s AI Action Plan (Dec. 22, 2025)). The 2024 order had barred Rytr from offering a service that generated consumer reviews and testimonials. Its vacatur signals that the FTC is less likely to pursue AI tool providers based on the potential for downstream misuse, but it does not affect the liability of businesses that actually use AI to create or disseminate fake reviews. On the same day it vacated the Rytr order, the FTC sent warning letters to companies regarding potential violations of the Consumer Reviews and Testimonials Rule (see Consumer Reviews and Testimonials Rule).

In July 2026, as directed by Executive Order 14365, the FTC issued a proposed policy statement on the application of Section 5 to AI systems. The proposed statement takes the position that AI companies that steer their systems’ outputs toward undisclosed objectives, rather than the objectives users set or reasonably expect, are likely to deceive consumers, and that state laws requiring such alterations may be preempted (FTC, Proposed Policy Statement Concerning the Suppression of Accuracy in Artificial Intelligence Systems (July 1, 2026)). The proposed statement is directed principally at AI developers rather than advertisers, but it confirms that the FTC views the deception framework as fully applicable to AI outputs.

The FTC has not adopted any rule or guidance requiring disclosure solely because GenAI was used to create or edit advertising content.

Endorsement Guides

The FTC’s Guides Concerning the Use of Endorsements and Testimonials in Advertising (Endorsement Guides) were most recently revised in 2023 (16 C.F.R. Part 255; 88 Fed. Reg. 48092 (July 26, 2023)). The Endorsement Guides are administrative interpretations of Section 5 rather than binding rules, but practices inconsistent with them may result in enforcement action (16 C.F.R. § 255.0(a)). The earlier article noted that the FTC treats fictional customer testimonials as inconsistent with the Endorsement Guides. The provisions most relevant to AI-generated influencer content are:

• Endorsers. An endorser is the party whose opinions, beliefs, findings, or experience a message appears to reflect, and it “could be or appear to be an individual, group, or institution” (16 C.F.R. § 255.0(b)). In announcing the 2023 revisions, the FTC explained that the definition reaches virtual influencers (FTC: Federal Trade Commission Announces Updated Advertising Guides to Combat Deceptive Reviews and Endorsements (June 29, 2023)).
• Honest opinions and experience. Endorsements must reflect the honest opinions, findings, beliefs, or experience of the endorser, and an endorsement may not convey any representation that would be deceptive if made directly by the advertiser (16 C.F.R. § 255.1(a)).
• Distortion. An endorsement may not be presented out of context or reworded in a way that distorts the endorser’s opinion or experience (16 C.F.R. § 255.1(b)).
• Current views. An advertiser may use an endorsement of an expert or celebrity only so long as it has good reason to believe that the endorser continues to subscribe to the views presented (16 C.F.R. § 255.1(b)).
• Bona fide use. If an advertisement represents that the endorser uses the product, the endorser must have been a bona fide user at the time of the endorsement, and the advertiser may continue running the advertisement only while it has good reason to believe the endorser remains a bona fide user (16 C.F.R. § 255.1(c)).
• Liability. Advertisers are liable for misleading or unsubstantiated statements made through endorsements and for failing to disclose unexpected material connections, and should provide guidance to endorsers, monitor compliance, and remedy non-compliance (16 C.F.R. § 255.1(d)). Endorsers may be liable for representations they know or should know to be deceptive, including falsely representing that they personally used a product (16 C.F.R. § 255.1(e)). Advertising agencies, public relations firms, and other intermediaries may be liable for their roles in creating or disseminating deceptive endorsements (16 C.F.R. § 255.1(f)).
• Images of other people. Using an image or likeness of someone other than the actual endorser is deceptive if it misrepresents a material attribute of the endorser (16 C.F.R. § 255.1(g)).
• Performance claims. Consumer endorsements about product performance must be supported by adequate substantiation, and consumer endorsements themselves are not competent and reliable scientific evidence (16 C.F.R. § 255.2(a)).
• Material connections. A connection between the endorser and the seller that might materially affect the weight or credibility of the endorsement, and that the audience would not reasonably expect, must be disclosed clearly and conspicuously (16 C.F.R. § 255.5(a)). In social media and other interactive media, the disclosure should be unavoidable (16 C.F.R. § 255.0(f)). A disclosure that appears only on an influencer’s profile page is not clear and conspicuous (16 C.F.R. § 255.0(g)(9)).

For FTC staff guidance on influencer disclosures, see FTC, Disclosures 101 for Social Media Influencers (Nov. 2019) and FTC, FTC’s Endorsement Guides: What People Are Asking (2023).

Consumer Reviews and Testimonials Rule

The FTC’s Trade Regulation Rule on the Use of Consumer Reviews and Testimonials took effect on October 21, 2024 (16 C.F.R. Part 465; 89 Fed. Reg. 68034 (Aug. 22, 2024)). The rule prohibits businesses from writing, creating, or selling consumer reviews, consumer testimonials, or celebrity testimonials that materially misrepresent:

• That the reviewer or testimonialist exists.
• That the reviewer or testimonialist used or otherwise had experience with the product.
• The reviewer’s or testimonialist’s experience with the product.

(16 C.F.R. § 465.2(a).) The rule also prohibits businesses from purchasing such reviews or disseminating such testimonials when they knew or should have known of the misrepresentation (16 C.F.R. § 465.2(b)). The rule’s text does not mention AI, but in announcing the rule the FTC specifically identified AI-generated fake reviews as within its scope (FTC: Federal Trade Commission Announces Final Rule Banning Fake Reviews and Testimonials (Aug. 14, 2024)).

Because it is a trade regulation rule, violations may result in civil penalties where the violator had actual knowledge, or knowledge fairly implied, that the conduct was prohibited by the rule (15 U.S.C. § 45(m)(1)(A)). Whether a particular influencer statement is a consumer testimonial or celebrity testimonial under the rule’s definitions requires case-by-case analysis (16 C.F.R. § 465.1). The rule is a meaningful addition to the framework discussed in the earlier article because it gives the FTC a direct civil penalty remedy for fabricated testimonials, including those generated by AI or delivered through a synthetic persona.

Impersonation Rule

The FTC’s Trade Regulation Rule on Impersonation of Government and Businesses took effect on April 1, 2024 (16 C.F.R. Part 461; 89 Fed. Reg. 15017 (Mar. 1, 2024)). The rule prohibits impersonating government entities, businesses, and their officials, and misrepresenting affiliation with or endorsement by them, but it does not reach impersonation of private individuals.

In 2024, the FTC proposed extending the rule to impersonation of individuals and adding means-and-instrumentalities liability that could reach providers of AI tools (89 Fed. Reg. 15072 (Mar. 1, 2024)). The FTC later decided not to proceed with the means-and-instrumentalities proposal. It held an informal hearing on the individual-impersonation proposal in January 2025 but, as of October 2026, had not finalized it. In October 2026, the FTC separately issued an advance notice of proposed rulemaking asking whether the rule should address platform advertising tools used in impersonation scams (FTC, Rule on Impersonation of Government and Businesses, Advance Notice of Proposed Rulemaking (Oct. 1, 2026)).

Even if the individual-impersonation provision is finalized, it would principally address unauthorized impersonation, such as deepfake celebrity endorsements, rather than an influencer-approved digital replica.

State UDAP Laws

UDAP statutes are in force in all states and the District of Columbia, provide broad prohibitions against unfair or deceptive conduct, and sometimes exceed federal standards. Many direct courts to be guided by FTC and federal court interpretations of Section 5 (see, for example, Mass. Gen. Laws ch. 93A, § 2(b); Fla. Stat. § 501.204(2)).

Utah has clarified that it is not a defense to a violation of a statute enforced by its Division of Consumer Protection that GenAI made the violative statement, undertook the violative act, or was used in furtherance of the violation. This provision, enacted in 2025, is codified at Utah Code § 13-77-102.

State attorneys general have long pursued fabricated endorsements and false authenticity signals under their general consumer protection authority. For example, in 2013 the New York Attorney General announced settlements with 19 companies following an investigation into fake online reviews, and in 2019 the New York Attorney General reached a settlement with Devumi LLC over its sale of fake social media followers and engagement, some of which used accounts copying real people’s profiles (see N.Y. Gen. Bus. Law §§ 349 to 350). These are traditional deception matters rather than AI-specific precedent, but they apply by analogy to AI-generated reviews and personas.

State AI Disclosure Laws

Several states have enacted AI-specific disclosure requirements relevant to influencer marketing.

New York Synthetic Performer Law

New York requires advertisements that include a synthetic performer to disclose that fact conspicuously (N.Y. Gen. Bus. Law § 396-b (signed Dec. 11, 2025; effective June 9, 2026)). A synthetic performer is a digitally created asset, generated or modified by AI or a software algorithm, that is intended to create the impression of a human performer who is not an identifiable real person. The disclosure obligation applies to persons who produce or create advertisements, and only where they have actual knowledge that a synthetic performer is used. The law includes exceptions, including for audio-only advertisements and certain uses in connection with expressive works, and does not apply to media outlets that merely carry advertising. Civil penalties are $1,000 for a first violation and $5,000 for each subsequent violation.

The New York law is the first state law to require disclosure of AI-generated performers in advertising and is directly relevant to brands that use fully synthetic influencers in campaigns reaching New York audiences.

California BOT Act

It is unlawful to use a bot to communicate or interact online with a person in California with the intent to mislead the person about its artificial identity in order to incentivize a purchase or sale of goods or services, unless the bot’s nature is clearly and conspicuously disclosed. A bot is an automated online account where all or substantially all of the actions or posts are not the result of a person (Cal. Bus. & Prof. Code §§ 17940 to 17943).

Utah GenAI Consumer Protection Law

Utah requires a supplier that uses GenAI to interact with an individual in connection with a consumer transaction to disclose that the individual is interacting with GenAI rather than a human, if the individual clearly and unambiguously asks. Individuals providing services in certain regulated occupations must prominently disclose GenAI use at the outset of high-risk AI interactions, such as those involving sensitive personal information or personalized financial, legal, medical, or mental health advice. The law provides a safe harbor for compliant disclosures (Utah Code §§ 13-77-101 to 13-77-106).

California AI Transparency Act

California requires providers of GenAI systems with more than one million monthly users to offer free AI detection tools and to include provenance disclosures in AI-generated image, video, and audio content (Cal. Bus. & Prof. Code § 22757 et seq.). The Act became operative on August 2, 2026, after AB 853 (2025) delayed its original January 1, 2026 operative date. AB 853 also adds provenance-related obligations for large online platforms and other entities beginning in 2027. The law regulates AI providers and platforms rather than advertisers, but AI-generated campaign assets may carry machine-readable provenance data that platforms can detect and label.

Federal Challenges to State AI Laws

These laws operate against an unsettled federal backdrop. Executive Order 14365 directs the Attorney General to establish an AI Litigation Task Force to challenge state AI laws deemed inconsistent with a minimally burdensome national AI policy framework, directs the Secretary of Commerce to identify such laws, and directs the FTC to issue the policy statement discussed above (Exec. Order No. 14365, Ensuring a National Policy Framework for Artificial Intelligence, 90 Fed. Reg. 58499 (Dec. 16, 2025)). The order does not itself invalidate any state law. Organizations should comply with state disclosure laws as enacted while monitoring for preemption challenges.

Rights of Publicity and Digital Replica Laws

State statutory and common law rights of publicity are the most relevant protections against unauthorized use of a person’s identity, but there is no federal right of publicity and state laws vary considerably. These laws restrict the unauthorized commercial use of a person’s name, likeness, or voice (see, for example, Cal. Civ. Code § 3344; N.Y. Civ. Rights Law §§ 50 to 51). Courts have long applied them to imitations that evoke a person’s identity without using the person’s actual image or voice (see, for example, Midler v. Ford Motor Co., 849 F.2d 460 (9th Cir. 1988) (deliberate imitation of a singer’s distinctive voice in a commercial)).

These protections are now being tested against AI voice cloning. In Lehrman v. Lovo, Inc., voice actors sued an AI voice company for creating and selling clones of their voices. The court dismissed the plaintiffs’ federal trademark claims and most of their copyright claims but allowed their claims under New York’s right of publicity statute, New York consumer protection law, and breach of contract to proceed (Lehrman v. Lovo, Inc., No. 1:23-cv-08269, 2025 WL 1902547 (S.D.N.Y. July 10, 2025)). The decision suggests that, at least for now, state publicity and consumer protection laws are more effective than federal intellectual property law against unauthorized AI voice replicas.

Several states have also enacted laws that specifically address digital replicas, including:

• Tennessee. The ELVIS Act extends publicity protection to an individual’s voice, including simulations of it, and creates liability for distributing technology whose primary purpose is producing a particular individual’s likeness or voice, with knowledge that the use is unauthorized (Ensuring Likeness, Voice, and Image Security (ELVIS) Act of 2024, 2024 Tenn. Pub. Acts ch. 588 (codified at Tenn. Code Ann. § 47-25-1101 et seq.) (eff. July 1, 2024)).
• California. Contract provisions permitting a digital replica of an individual’s voice or likeness in place of work the individual would otherwise have performed in person may be unenforceable where they lack a reasonably specific description of the intended uses and the individual was not represented by counsel or a labor union (Cal. Lab. Code § 927 (applicable to performances fixed on or after Jan. 1, 2025)). Commentators disagree whether a reasonably specific description or representation alone is sufficient to preserve a provision, so agreements should satisfy both. California also requires consent for digital replicas of deceased personalities in audiovisual works and sound recordings (Cal. Civ. Code § 3344.1).
• New York. New York imposes comparable limits on digital replica contract provisions (N.Y. Gen. Oblig. Law § 5-302 (applicable to contracts entered into or modified on or after Jan. 1, 2025)). New York also requires consent before a deceased performer’s digital replica is used. In December 2025, New York broadened that protection to cover digital replicas used in audiovisual works, sound recordings, and live musical performances (N.Y. Civ. Rights Law § 50-f (as amended Dec. 11, 2025)).
• Illinois. Illinois amended its Right of Publicity Act in 2024 to address unauthorized digital replicas (765 ILCS 1075) and enacted the Digital Voice and Likeness Protection Act, which limits the enforceability of digital replica contract provisions (815 ILCS 550/1 et seq.).
• Other states. Utah, Arkansas, Montana, and Washington have also enacted legislation addressing AI-generated digital replicas of a person’s voice or likeness.

An influencer whose replica is used without authorization may also have a federal false endorsement claim under the Lanham Act (15 U.S.C. § 1125(a)(1)(A); Waits v. Frito-Lay, Inc., 978 F.2d 1093 (9th Cir. 1992)).

The No FAKES Act, which would create a licensable federal right in a person’s voice and visual likeness in digital replicas, was reintroduced in the 119th Congress as the NO FAKES Act of 2026, which adds a counter-notification process to its notice-and-takedown regime and would preserve state digital replica laws in existence as of January 2, 2025. The Senate Judiciary Committee approved the bill in June 2026, and as of October 2026 it remains pending on the Senate calendar (S. 4591, 119th Cong. (2026)). The TAKE IT DOWN Act, enacted in May 2025, addresses nonconsensual intimate imagery, including AI-generated imagery, rather than advertising (Pub. L. No. 119-12 (2025)).

These laws generally protect an individual’s control over their identity. They do not resolve whether consumers are being misled, which remains the central question under advertising law.

Platform Policies

Major social media platforms require creators to label realistic AI-generated or materially altered content. For example:

• YouTube requires creators to disclose realistic altered or synthetic content, but does not require disclosure where GenAI is used only for productivity tasks such as drafting scripts or captions.
• TikTok requires creators to label AI-generated content that contains realistic images, audio, or video.
• Meta applies “AI info” labels and requires disclosure of photorealistic video or realistic-sounding audio that was digitally created or altered.

These requirements are contractual terms rather than legal mandates, but noncompliance can result in content removal or account penalties. Platform policies change frequently and should be checked for each campaign.

International Laws

The EU AI Act has extraterritorial reach and may apply to organizations outside the EU that produce AI system output used, or intended to be used, in the EU (Regulation (EU) 2024/1689 (EU AI Act)). Its transparency obligations require deployers of AI systems that generate or manipulate deepfake image, audio, or video content to disclose that the content has been artificially generated or manipulated (Article 50(4), EU AI Act).

Article 50 has applied since August 2, 2026 (Article 113, EU AI Act). The Digital Omnibus on AI, adopted in 2026, deferred the EU AI Act’s high-risk system obligations but did not defer Article 50, except for a grace period until December 2, 2026 for the Article 50(2) marking obligations of GenAI systems placed on the market before August 2, 2026. Campaigns that reach EU audiences and include AI replicas of real people, or synthetic content that could be mistaken for authentic, should account for these labeling requirements.

Applying the Framework

The legal significance of AI in influencer marketing depends less on whether AI was used than on what the AI does. Under federal advertising law, the key questions are whether AI creates or simulates a real person’s statements or experience, creates a fictional persona, or materially alters a product demonstration, and whether consumers would understand the advertisement differently if they knew. Separately, state, platform, and foreign rules may require AI disclosure regardless of the deception analysis.

Conclusion 

In the first article in this series, “AI, Influencers, and Liability: What Marketing Agencies Need to Know Before Deploying AI Tools,” we explained that the key legal question for brands and agencies is not simply whether AI is being used in influencer marketing, but how it is being used. The developments discussed in this article reinforce that point. Federal law still does not require disclosure solely because AI was used, and the FTC has signaled that it will apply existing deception and substantiation principles to AI rather than create AI-specific obligations. Those principles apply fully to AI-generated content: endorsements must reflect genuine opinions and experiences, claims must be substantiated, and the Consumer Reviews and Testimonials Rule now gives the FTC a civil penalty remedy for fabricated testimonials, including those generated by AI.

At the same time, AI-specific obligations are emerging outside federal advertising law. New York now requires disclosure of synthetic performers in advertising, states continue to expand protections for digital replicas, platforms require labels for realistic AI content, and the EU AI Act’s deepfake disclosure requirements now apply. These rules operate against an unsettled federal backdrop, including Executive Order 14365’s challenge to state AI laws and the pending NO FAKES Act, so the requirements that apply to a given campaign may continue to shift.

Brands and agencies should comply with state disclosure laws as enacted, track platform and international requirements by campaign and audience, address AI use and digital replica rights expressly in influencer agreements, and monitor federal preemption efforts. Those that evaluate AI based on what it does, rather than simply whether it was used, will be better positioned to use GenAI effectively while minimizing legal and reputational risk.

Our next article, “AI-Assisted, Replicated, and Synthetic Influencers: Legal Risks and When to Disclose AI,” takes a closer look at each of the three ways AI is used in influencer marketing and offers a practical framework for deciding when AI use should be disclosed.

For further information and assistance, contact Lindsay Gehman or Saachi Gorinstein.

To view a PDF version of this article, please click here.