FTC Group Letter: Call for FTC Oversight, Investigation, and Suspension of Meta’s AI Chatbot Advertising Practice
October 30, 2025
Andrew Ferguson
Chair
Federal Trade Commission
600 Pennsylvania Ave NW
Washington, DC 20580
Re. Call for FTC Oversight, Investigation, and Suspension of Meta’s AI Chatbot Advertising Practice
Dear Chair Ferguson, Commissioner Holyoak, and Commissioner Meador:
We, the undersigned 36 organizations dedicated to privacy, data justice, civil rights, children’s rights, consumer protection, and research, write to demand immediate action in response to Meta Platforms, Inc.’s October 1, 2025, announcement that, effective December 16, 2025, it will begin using user interactions with AI chatbots (voice and text) to personalize advertising and content.[1]
This unprecedented use of deeply sensitive data presents outsized risks to consumer data privacy and security, and is at odds with the FTC’s 2020 Order secured against Facebook (now Meta). It also underscores the urgent need for enforcement of Section 5 of the FTC Act against unfair and deceptive AI practices. The danger of regulatory inaction at this critical moment, when generative AI is driving an unprecedented expansion of commercial surveillance, could be catastrophic and irreversible.
Meta’s Expanding Reliance on Generative AI for Surveillance Marketing
The Commission should recognize that Meta’s decision to monetize AI-driven chatbots without even a basic opt-out mechanism reflects the company’s broader strategy: an aggressive expansion of AI for marketing and advertising. More than half of all U.S. internet users—163 million people—are projected to use generative AI by 2029, making it one of the fastest-adopted technologies in modern history, outpacing both personal computers and the internet in their early phases.[2] Roughly one third of Americans under 30 already engage with AI several times daily, and it is projected that by 2029, more than 55 million Gen Z users (ages 18–34) will rely on these tools.[3] Meta is at the center of this transformation. Its platforms, Facebook and Instagram, remain dominant among U.S. internet users, and its advertising models are increasingly driven by AI systems such as Advantage+ shopping and ad services, “audience optimization,” automated creative testing, video personalization, conversational search, and virtual influencers.[4] WARC recently described Meta’s AI ad system as enabling marketers to leverage “the richest possible signals to train AI with,” underscoring that the company’s marketing innovations depend on harvesting ever greater amounts of user data.[5]
Meta’s harvesting of AI chatbot interactions for advertising must be seen in this broader context. It is not a marginal feature but rather part of a deliberate strategy to normalize a fundamental expansion of surveillance-driven and behavior-changing marketing. The company’s AI-optimized formats are explicitly designed to drive “actionable” reach by shaping how people spend their money, time, and attention.[6] Analysts have already suggested that Meta’s chatbot could “dramatically boost engagement” by initiating unprompted conversations based on prior interactions.[7] Even eMarketer, an industry research group, cautioned in response to Meta’s October 1, 2025 announcement that “guardrails on ad placement within chatbot conversations need to be tight to prevent what could be perceived as exploitive targeting.”[8] These developments carry serious consequences for all users—especially youth—by embedding invasive AI data practices into daily online interactions without meaningful safeguards. They also underscore why the Commission must exercise its oversight authority under the consent decree and, if necessary, pursue enforcement under Section 5 of the FTC Act to ensure Meta’s practices are lawful, transparent, and subject to appropriate safeguards.
Consent Decree Obligations Require FTC Oversight and Investigation
The FTC’s 2020 Order[9] against Meta (formerly Facebook) imposes clear obligations on Meta, and the FTC has both authority and a responsibility to conduct oversight to ensure these obligations are being met. It is crucial that the FTC exercise this oversight authority here.
Parts VII and VIII of the 2020 Order require that Meta, among other things:
Conduct a detailed product review or data privacy and security risk assessment for each new or modified product, service, or practice that collects or uses consumer data;
Document and implement safeguards to protect consumer privacy in response to any risks identified;
Retain records for FTC or independent assessor review; and
Obtain an independent biennial assessment of its program and produce a Quarterly Privacy Review Report, which must be furnished to the Commission upon request.
Meta’s proposed use of conversational AI data for advertising and content personalization is plainly a “new or modified” practice under this framework. Conversational data generated through AI chat interactions is substantially more sensitive than ordinary behavioral data, as it may reveal personal relationships, mental health concerns, political views, and other intimate information.[10] Accordingly, Meta must conduct a comprehensive privacy and security risk assessment of this practice, and that assessment must be subject to formal FTC oversight, review, and documentation.
Furthermore, under the modified order, Meta is prohibited from misrepresenting the extent to which it maintains the privacy or confidentiality of user information.[11] While Meta announced the launch of generative AI features across its apps on September 27, 2023, it did not disclose until October 1, 2025, that it would collect and use AI-chat interactions for content and ad personalization and targeting.[12] That announcement was followed by an email to users with only a brief update to the Privacy Policy: “We’ll start using your interactions with AIs to personalize your experiences and ads.” Most users do not follow corporate blogs, and failing to clearly address that AI-chat interactions are covered, or to provide meaningful choices such as an opt-in, suggests an effort to obscure the risks of this practice. This lack of transparency requires investigation under the consent decree’s provisions. Without clear and forthright disclosure, the public has no assurance that Meta’s practices comply with its obligations or that the safeguards envisioned by the decree are functioning as intended.
We call on the Commission to exercise its oversight authority under the consent decree and publicly disclose its findings.
The FTC Must Finalize the 2023 Proposed Modifications to the 2020 Order
We are also deeply concerned that the Commission has stayed proceedings on the 2023 proposed modifications to the 2020 Order, which would strengthen existing obligations and prohibit Meta from monetizing minors’ data except for limited, security-related purposes. This delay is indefensible as Meta escalates its commercial surveillance practices.
By failing to finalize these safeguards, the Commission leaves children and teenagers—those most at risk of manipulation through AI-driven advertising—without meaningful protection. It also signals to Meta and the broader industry that there are no enforceable limits on the use of conversational data, AI-derived inferences, or other sensitive digital interactions. This passivity undermines the Commission’s credibility and exposes consumers to escalating risks as generative AI expands the scope and scale of commercial surveillance.
The FTC Must Initiate a Section 5 Investigation and Meta Must Suspend Implementation
Even beyond the consent decrees, Meta’s proposal appears to constitute an unfair or deceptive practice in violation of Section 5 of the FTC Act.
· Deceptive practice: Consumers reasonably expect that private chats and conversational interactions with digital assistants will not be monetized for advertising unless explicitly and clearly disclosed. Meta’s plan to use chatbot interactions for ad targeting and personalization, without prior clear disclosure and opt-in consent, is a material omission or misrepresentation likely to mislead consumers.
· Unfair practice: The use of private conversational data for advertising imposes a substantial injury that consumers cannot reasonably avoid, and which is not outweighed by countervailing benefits. Harms include loss of privacy, the chilling of free expression, discrimination via algorithmic profiling, and reputational or psychological risk. Minors, in particular, may be harmed by personalized social media feeds whose content is optimized for engagement, producing a variety of downstream harms such as exposure to harmful content, reinforcement of unhealthy behaviors, and adverse effects on mental health.
The Commission should open a Section 5 investigation and direct Meta to suspend implementation of this practice unless and until the Commission has completed a public investigation and determined that it complies with federal law.
FTC Leadership and Decisive Action Needed Now
The FTC has repeatedly affirmed that the deployment of artificial intelligence does not exempt firms from existing consumer protection obligations. Yet Meta’s plan to repurpose conversational AI data for advertising illustrates the dangers of regulatory inaction: absent intervention, the practice will normalize a level of surveillance that is qualitatively more intrusive than traditional behavioral tracking.
Conversational data is uniquely revealing, capturing intimate details of users’ personal lives, relationships, health, and beliefs. It therefore warrants heightened scrutiny under both the consent decrees and Section 5 of the FTC Act. At a minimum, this category of data should be treated as sensitive information requiring affirmative, opt-in consent from adults (and a categorical prohibition on monetization of youth data) before it can be used for advertising. Failing to impose this standard would effectively permit firms to unilaterally exploit the most private forms of digital interaction for commercial gain.
Requests for Relief
Accordingly, we respectfully but urgently request that the Commission do the following:
1. Publicly confirm whether it has reviewed Meta’s privacy and security risk assessments of Meta’s proposed new practice to use AI chat personal data for ad personalization and targeting. The FTC must address whether Meta has engaged in any misrepresentations regarding these new practices. The Commission must release its findings to the public.
2. Initiate a formal Section 5 investigation into Meta’s unfair and deceptive use of chatbot data for advertising and content personalization without informed, opt-in consent.
3. Require Meta to suspend implementation of this practice (slated to go into effect December 16, 2025) pending completion of the Commission’s investigations, public disclosure of findings, and approval under both the FTC’s consent-decree oversight and enforcement authority.
4. Immediately resume and finalize the 2023 proposed amendments to the 2020 Order, particularly those prohibiting monetization of minors’ data and other high-risk data types, without further delay.
5. Issue public guidance reaffirming that firms are not exempt from privacy obligations under the FTC Act when they use AI and generative AI and clarifying that conversational interaction data is presumptively sensitive, warranting opt-in consent for adults and categorical prohibition for minors.
The Commission has both the authority and the responsibility to act. Failure to do so will permit Meta to proceed unchecked and embolden other companies to adopt similarly invasive AI-driven practices, eroding consumer trust and exposing millions of Americans, especially children and vulnerable communities, to unbounded digital exploitation.[13]
We urge you to act without delay. If you would like to discuss further, please contact Jeff Chester (jeff@democraticmedia.org) or John Davisson (davisson@epic.org).
Respectfully,
Center for Digital Democracy
Electronic Privacy Information Center (EPIC)
20/20 Vision
5Rights Foundation
Arkansas Community Organizations
Becca Schmill Foundation
Blue Rising
Center for Economic Integrity
Center for Oil & Gas Organizing
Check My Ads Institute
Children and Screens: Institute of Digital Media and Childhood Development
Children Now
Common Sense Media
Consumer Action
Consumer Federation of America
Demand Progress Education Fund
Equality New Mexico
Fairplay
Indigenous Women Rising
Investor Alliance for Human Rights
Lynn’s Warriors
Mossville Environmental Action Now (MEAN)
Mothers Against Media Addiction (MAMA)
National Association of Consumer Advocates
New Jersey Appleseed Public Interest Law Center
Open Media and Information Companies Initiative (Open MIC)
Oregon Consumer Justice
Oregon Consumer League
ParentsSOS
ParentsTogether Action
Public Citizen
Suicide Awareness Voices of Education (SAVE)
Tech Justice Law Project
Tech Oversight Project
UltraViolet
Virginia Citizens Consumer Council
[1] Improving Your Recommendations on Our Apps With AI at Meta, Meta (Oct. 1, 2025), https://about.fb.com/news/2025/10/improving-your-recommendations-apps-ai-meta/.
[2] Jacob Bourne, GenAI User Forecast 2025, eMarketer (July 22, 2025), https://content-na1.emarketer.com/genai-user-forecast-2025.
[3] Id.
[4] Jeremy Goldman, Instagram, Facebook Help Meta Gain Share of US Social Ad Spending for First Time Since 2018, eMarketer (Oct. 13, 2025), https://content-na1.emarketer.com/meta-reclaims-ad-spending-momentum-even-reddit-breaks.
[5] WARC, What We Know About Marketing on Facebook (Aug. 2025), www.warc.com/SubscriberContent/article/bestprac/what-we-know-about-marketing-on-facebook/en-gb/110336.
[6] Id.
[7] Bourne, supra note 2.
[8] Jacob Bourne, AI, Consumer Behavior, and the Trust Economy, eMarketer (Oct. 20, 2025), https://content-na1.emarketer.com/ai-consumer-behavior-trust-economy.
[9] Modified Decision and Order, In the Matter of Facebook, Inc., C-4365 (FTC Apr. 27, 2020), www.ftc.gov/system/files/documents/cases/c4365facebookmodifyingorder.pdf.
[10] Empirical research, teen surveys, and litigation show that users often disclose highly sensitive personal information to AI chatbots, including names, locations, secrets, mental health crises, and sexual preferences. Design choices (friend-like personas, nudges, and non-judgmental tones) help explain why generative chatbots elicit such intimate disclosures. See Common Sense Media, Talk, Trust & Trade-Offs: How and Why Teens Use AI Companions (July 16, 2025), www.commonsensemedia.org/research/talk-trust-and-trade-offs-how-and-why-teens-use-ai-companions; see also Kate Payne, An AI Chatbot Pushed a Teen to Kill Himself, a Lawsuit Alleges, Associated Press (Oct. 25, 2024), www.apnews.com/article/chatbot-ai-lawsuit-suicide-teen-artificial-intelligence-9d48adc572100822fdbc3c90d1456bd0; Niloofar Mireshghallah et al., Trust No Bot: Discovering Personal Disclosures in Human-LLM Conversations in the Wild, arXiv e-prints (July 2024), https://arxiv.org/abs/2407.11438; Naomi Nix & Nitasha Tiku, Meta AI Users Confide on Sex, God and Trump. Some Don’t Know It’s Public, Washington Post (June 13, 2025), www.washingtonpost.com/technology/2025/06/13/meta-ai-privacy-users-chatbot/.
The FTC recently noted that chatbots “are designed to communicate like a friend or confidant, which may prompt some users, especially children and teens, to trust and form relationships with chatbots.” Federal Trade Commission, FTC Launches Inquiry into AI Chatbots Acting as Companions (Sept. 2025), www.ftc.gov/news-events/news/press-releases/2025/09/ftc-launches-inquiry-ai-chatbots-acting-companions.
[11] See Modified Decision and Order, In the Matter of Facebook, Inc., C-4365, at 5 (FTC Apr. 27, 2020), www.ftc.gov/system/files/documents/cases/c4365facebookmodifyingorder.pdf.
[12] Meta announced the launch of its AI assistants for consumers on September 27, 2023, accompanied by the post “Privacy Matters: Meta’s Generative AI Features.” See Mike Clark, Privacy Matters: Meta’s Generative AI Features, Meta (Sep. 27, 2023), https://about.fb.com/news/2023/09/privacy-matters-metas-generative-ai-features/. Public statements by Meta about its use of chatbot or AI-assistant interaction data from U.S. users—as distinct from publicly available posts or shared content—are difficult to identify with precision. A disclosure on “artificial intelligence technology” linked to Meta’s June 26, 2024 privacy policy, however, states: “This is why we draw from publicly available and licensed sources, as well as information people have shared on Meta Products, including interactions with AI at Meta features. We keep training data for as long as we need it on a case-by-case basis to ensure an AI model is operating appropriately, safely and efficiently.” Meta, How Meta Uses Information for Generative AI Models and Features (last visited Oct. 29. 2025), https://www.facebook.com/privacy/genai.
This and subsequent corporate disclosures, see, e.g., Meta, Meta’s AI Products Just Got Smarter and More Useful (Sep. 25, 2024), https://about.fb.com/news/2024/09/metas-ai-product-news-connect/; Meta, Building Toward a Smarter, More Personalized Assistant (Jan. 27, 2025), https://about.fb.com/news/2025/01/building-toward-a-smarter-more-personalized-assistant/, reference the use of chatbot interaction data for training but do not identify advertising or content personalization as a purpose. Meta did not announce such uses until October 1, 2025, when it stated that chatbot interactions would be leveraged for ad targeting and content personalization beginning December 16, 2025.
[13] Laurie Sullivan, OpenAI Browser Keeps Search, Adds Memory Benefiting Advertisers, MediaPost (Oct. 21, 2025), www.mediapost.com/publications/article/410064/openai-browser-keeps-search-adds-memory-benefitin.html.