Comments from the Electronic Privacy Information Center (EPIC) Responding to the Request for Information for Regulatory Reform on Artificial Intelligence

To the Office of Science and Technology Policy:

The Electronic Privacy Information Center (EPIC) submits these comments in response to the Office of Science and Technology Policy (OSTP)’s Request for Information (RFI) to identify “existing Federal statutes, regulations, agency rules, guidance, forms, and administrative processes that unnecessarily hinder the development, deployment, and adoption of artificial intelligence (AI) technologies within the United States.”[1] We reject the framing that existing laws act as a hindrance to AI rather than a necessary protection for those affected by it. Innovation and regulation need not be at odds – indeed, the supplementary information provided with the RFI notes that “[s]uitable policy frameworks enable innovation while safeguarding the public interest.” However, the RFI itself fails to inquire about what regulations are needed to ensure that AI systems are developed responsibly, instead asking only for which laws hinder its development, deployment, and adoption. The federal government has long regulated product, drug, and airline safety, among other areas, to protect consumers. AI should not get a special pass from this tried-and-true consumer protection. Americans have repeatedly called for more regulation on AI rather than less. We call on the OSTP to withdraw this dangerous inquiry.

EPIC is a public interest research center in Washington, D.C., established in 1994 to focus public attention on emerging civil liberties issues and to secure the fundamental right to privacy in the digital age for all people through advocacy, research, and litigation.[2] EPIC has worked with various federal and state agencies, including the OSTP, to strengthen privacy protections and AI safeguards.[3] 

This RFI Ignores Existing AI Harms

The RFI acknowledges that AI systems are already deployed across various sectors of the economy and public life. However, the framing ignores the reality that this rapid deployment of AI systems has resulted in multiple severe harms to Americans and that existing regulations are insufficient to fully address such harms. Many forms of AI, including generative AI and automated decision making systems (ADS), produce opaque, inaccurate, or biased outputs in consequential decisions such as housing, employment, education, government benefits, healthcare, and criminal justice.[4] AI systems are often unaccountable in their decision making process—both to enforcement bodies left unable to analyze and explain how data factors are weighed in generated recommendations or decisions, but also to AI creators unable to do so. As nontransparent, unaccountable AI systems persistently issue decisions resulting in discriminatory, damaging impacts, they threaten to erode hard-won antidiscrimination laws and civil rights under the false guise of an “objective” system.

Businesses also use AI for surveillance pricing and sharing pricing data in violation of antitrust laws to increase living costs when average Americans are struggling to make ends meet.[5] Generative AI produces a litany of harms, including collecting training data without consent, scams, hallucinations, disinformation, deepfake porn, false claims regarding qualifications and safety, mental health harms, and endangering elections.[6] AI companion chatbot threats to children have captured public attention due to multiple tragic chatbot-fueled teenager deaths by suicide.[7] The American public is aware of, and increasingly weary of, AI’s negative impact in consequential decisions, chatbot interactions, and more.[8]

AI Is Harming Vulnerable Groups

The OSTP must not disregard the current and ongoing harms caused by the use of AI. AI harms disproportionately impact vulnerable and marginalized groups. For example, companion AI chatbots are more harmful to children, who are more vulnerable to inappropriate sexual content and fabricated intimacy; AI systems used to determine criminal sentencing, surveil immigrants, determine housing eligibility, and more disproportionately impact low-income individuals and people of color. These harms have been well-documented over the last two decades and must be addressed.

While the RFI correctly acknowledges that many of the existing laws and regulations that affect AI development and deployment were passed without directly considering AI, that is an argument for more fulsome protections, not less. The RFI fails to inquire about or consider existing regulatory gaps and what new regulations or modifications are needed to ensure AI development is trustworthy, safe, and secure for the American public. A focus solely on deregulation rather than common-sense regulation benefits large corporate interests only—not the American people.

Deregulation Threatens Stability

Suggesting that a federal agency intends to abuse its power to undermine the existing rule of law in favor of an already harmful technology is not only unethical and inappropriate but will actively harm both the American people and the overall AI market. Removing the few current AI protections effectively punishes any businesses attempting to enter the AI space responsibly. A business operating legally and safely will dedicate resources to reviewing legal obligations, develop their project in safe ways rather than the fastest and easiest ways, and avoid risky shortcuts. Retroactively changing the playing field makes all these efforts a waste of time and rewards AI companies that developed unsafely and illegally from the start.

Deregulation threatens public expectations and safety. Previous efforts to block AI regulation have been roundly rejected and criticized by the American public, as in the attempted federal moratorium on state AI regulation. Targeting existing regulations for removal or modification communicates to the American people that their rights and safety are secondary to business interests. This will not inspire more trust in AI, only less trust in authorities meant to protect us.

The deregulatory agenda that underscores OSTP’s RFI threatens the effort to protect the privacy, autonomy, and civil rights of individuals. Undermining those protections would amount to abdication of responsibility to the needs of the American public. We strongly urge the OSTP to withdraw this RFI and focus on what regulatory actions are necessary to ensure that any AI development, deployment, and adoption advances the public interest.

[1] Office of Science and Technology Policy, Request for Information, Regulatory Reform on Artificial Intelligence, 90 Fed. Reg. 46422 (Sept. 26, 2025).

[2] EPIC, About Us (2025), https://epic.org/about/.

[3] Comments of EPIC to the White House OSTP, Request for Information: Automated Worker Surveillance and Management (June 15, 2023), https://epic.org/documents/epic-comments-to-white-house-ostp-on-automated-worker-surveillance/; EPIC, FTC Rulemaking on Commercial Surveillance & Data Security (2022), https://epic.org/ftc-rulemaking-on-commercial-surveillance-data-security/; EPIC, Risk Assessments (2025), https://epic.org/issues/ai/risk-assessments/.

[4] EPIC, “Generating Harms: Generative AI’s Impact & Paths Forward” (2023), https://epic.org/generating-harms/; EPIC, “Generating Harms II: Generative AI’s New & Continued Impacts” (2024), https://epic.org/generating-harms/; Mayu Tobin-Miyaji, “Assessing the Assessments: Maximizing the Effectiveness of Algorithmic & Privacy Risk Assessments Part I,” EPIC (2025), https://epic.org/wp-content/uploads/2025/06/Assessing-the-Assessments-Report.pdf.

[5] FTC Surveillance Pricing 6(b) Study: Research Summaries (Jan. 2025), https://www.ftc.gov/system/files/ftc_gov/pdf/p246202_surveillancepricing6bstudy_researchsummaries_redacted.pdf; Mayu Tobin-Miyaji, Kroger’s Surveillance Pricing Harms Consumers and Raises Prices, With or Without Facial Recognition, EPIC (Feb. 14, 2025), https://epic.org/krogers-surveillance-pricing-harms-consumers-and-raises-prices-with-or-without-facial-recognition/; Jay Stanley, “Surveillance Pricing” Hurts Consumers, Incentivizes More Corporate Spying on Them, ACLU (Sept. 12, 2025), https://www.aclu.org/news/privacy-technology/surveillance-pricing.

[6] EPIC, supra Note 4.

[7] Id; Cristiano Lima-Strong, Transcript: US Senate Hearing On ‘Examining the Harm of AI Chatbots’, Tech Policy Press (Sept. 17, 2025), https://www.techpolicy.press/transcript-us-senate-hearing-on-examining-the-harm-of-ai-chatbots/; Samantha Cole, Instagram’s AI Chatbots Lie About Being Licensed Therapists, 404 Media (April 29, 2025), https://www.404media.co/instagram-ai-studio-therapy-chatbots-lie-about-being-licensed-therapists/.

[8] Grace Gedye, “Consumer Reports Survey: Many Americans Concerned About AI, Algorithms,” Consumer Reports (July 25, 2024), https://advocacy.consumerreports.org/press_release/consumer-reports-survey-many-americans-concerned-about-ai-algorithms/; Elizabeth Laird, Maddy Dwyer & Hannah Quay-de la Vallee, “Hand in Hand: Schools’ Embrace of AI Connected to Increased Risk to Students,” CDT (2025), https://cdt.org/wp-content/uploads/2025/10/FINAL-CDT-2025-Hand-in-Hand-Polling-100225-accessible.pdf; Common Sense Media, “Talk, Trust, and Trade-Offs: How and Why Teens Use AI Companions” (2025), https://www.commonsensemedia.org/sites/default/files/research/report/talk-trust-and-trade-offs_2025_web.pdf.