⚡NIMITZ TECH NEWS FLASH⚡

“AI Regulation and the Future of US Leadership”

House Committee on Energy and Commerce, Subcommittee on Commerce, Manufacturing, and Trade”

May 21, 2025 (recording linked here)

HEARING INFORMATION

Witnesses and Written Testimony (Linked):

• Mr. Sean Heather: Senior Vice President, U.S. Chamber of Commerce

• Mr. Adam Thierer: Senior Fellow, R Street Institute

• Mr. Marc Bhargava: Managing Director, General Catalyst

• Ms. Amba Kak: Co-Executive Director, AI Now Institute
  

HEARING HIGHLIGHTS

IN THEIR WORDS

SUMMARY OF OPENING STATEMENTS FROM THE COMMITTEE AND SUBCOMMITTEE

• Subcommittee Chair Bilirakis emphasized that AI is advancing rapidly and holds transformative potential for the economy, but warned that it also presents serious risks if misused. Bilirakis argued that while existing laws provide some safeguards, overly burdensome new regulations could hinder innovation, especially among startups, and ultimately jeopardize U.S. leadership in the global AI race.

• Subcommittee Ranking Member Schakowsky stressed that her top concern was consumer protection and expressed outrage at proposals to allow a ten-year delay in accountability for AI-related harms. Schakowsky described this idea as re kless and warned that consumers are already being scammed due to inadequate oversight. She called for immediate safeguards and rejected the notion of deferring action for a decade.

• Rep. Trahan sharply criticized Republican leadership for failing to protect Americans’ privacy and children online. She condemned their opposition to whistleblower protections, data minimization, and transparency legislation, while pushing through a 10-year ban on state AI safety laws. Harshbarger argued that this move benefits Big Tech CEOs aligned with Trump, not everyday families. She called on her colleagues to reject the measure and instead pursue bipartisan action to hold tech companies accountable.

• Full Committee Vice Chair Joyce acknowledged the transformative potential of AI and emphasized the need for thoughtful regulation that supports innovation. He warned that a patchwork of state laws and the EU’s overregulation threaten to stifle American competitiveness, particularly for startups. Joyce advocated for a unified national framework that balances consumer protection with growth. He concluded by supporting targeted, narrowly tailored legislation and reaffirmed the committee’s commitment to American leadership in AI development.

• Rep. Schrier, speaking on behalf of Ranking Member Pallone, condemned a Republican-supported provision that bans state AI consumer protection laws for ten years. She emphasized the urgency of passing strong federal data privacy legislation, including safeguards against misuse of personal data by AI systems. Schrier highlighted that existing state laws already address real harms like deepfakes and discriminatory algorithmic decisions, which would be nullified under the proposed ban. She warned that Republicans are undermining both state efforts and federal enforcement capacity, favoring Big Tech over American consumers.

SUMMARY OF WITNESS STATEMENT

• Mr. Heather stated that the U.S. Chamber of Commerce strongly believes excessive AI regulation—especially the EU's approach—threatens innovation and competitiveness. He warned that the EU AI Act imposes unnecessary burdens without first reviewing existing legal frameworks, and that its code of practice could force companies to reveal sensitive intellectual property. Heather stressed that this would hurt U.S. firms globally, with the EU’s regulatory model already influencing states like California and Colorado. He concluded that to maintain AI leadership, the U.S. must foster a regulatory environment that supports innovation and avoids duplicating Europe's mistakes.

• Ms. Kak emphasized that AI leadership must deliver tangible benefits to the American public, not just to tech firms. She cited real-world harms caused by unregulated AI, including child suicide linked to chatbots and scams targeting seniors with voice cloning. Kak criticized Big Tech's dominance and the proposed moratorium on state AI laws, arguing that states have led effectively with targeted, bipartisan measures. She urged Congress not to shield powerful firms from oversight while ordinary Americans remain exposed to harm.

• Mr. Thierer warned that American AI innovation is being squeezed between Europe's overregulation and a fragmented U.S. state-by-state approach. He advocated for a federal moratorium on state AI laws, arguing that burdensome and inconsistent state regulations would crush startups and hurt competitiveness. Thierer cited existing legal tools that already address AI harms and noted bipartisan support, including from governors, for limiting regulatory overreach. He urged Congress to craft a national framework that supports interstate AI development while maintaining existing consumer protections.

• Mr. Bhargava described how General Catalyst invests in AI startups and emphasized the need for a balanced regulatory approach that ensures safety and rights without stifling innovation. He warned that overly rigid or premature regulation could drive innovation offshore, while the absence of guardrails risks societal harm. Bhargava recommended a national framework, ethical diligence in investing, and deeper collaboration between government, industry, and civil society. He urged Congress to emulate past successes like the Telecommunications Act of 1996 and adopt flexible standards such as model cards, audits, and public R&D investment.

SUMMARY OF KEY Q&A

• Chairman Bilirakis asked whether the EU was unfairly targeting American companies with enforcement actions under GDPR and if the same pattern might emerge under the AI Act. Mr. Heather responded that Europe has disproportionately fined American firms and embedded this enforcement approach into multiple laws, including the AI Act, warning of continued regulatory bias.

  Chairman Bilirakis then asked how venture capital firms are addressing AI risks and how government can support innovation while protecting consumers. Mr. Bhargava explained that General Catalyst evaluates training data, model outputs, risk mitigation strategies, and potential harms before investing, and encouraged the federal government to establish clear AI standards.

• Ranking Member Schakowsky asked about which AI harms pose the most urgent consumer protection threats and should be regulated. Ms. Kak warned that predatory AI tools target vulnerable users, secretive algorithms distort economic opportunity, and without current legal protections, consumers and small businesses are left defenseless against harm.

• Vice Chair Fulcher asked how GDPR and similar European laws have impacted U.S. companies trying to export to Europe. Mr. Heather stated that European regulations like GDPR have harmed U.S. competitiveness, especially for small businesses, and predicted the AI Act would replicate these trade barriers.

  The Vice Chair Fulcher then if the U.S. still leads in AI and how it can maintain that edge over China. Mr. Bhargava replied that the U.S. maintains a narrow lead due to its startup ecosystem, but preserving that advantage requires national guidelines rather than a fragmented patchwork of state regulations.

• Rep. Castor criticized the 10-year AI regulatory moratorium and cited a Wall Street Journal exposé on harmful AI chatbot interactions with children, then asked Ms. Kak how profitable this model is for tech companies. Ms. Kak said the AI industry, especially around children’s data, operates on a harmful business model centered on surveillance and engagement, and called for strong regulations to curb exploitative practices.

• Rep. Harshbarger asked whether the U.S. should pursue a voluntary federal trust label for AI and what elements it should include. Mr. Thierer responded that while there are promising national ideas, any approach must avoid Europe’s rigid, top-down model in favor of a flexible, bottom-up U.S. framework. Rep. Harshbarger then asked whether U.S. AI developers can expect regulatory certainty similar to China’s exemptions for non-public-facing AI work. Mr. Thierer replied that China’s model is based on surveillance and control, and emphasized that the U.S. must get its policy right to uphold its democratic values and global leadership.

  Rep. Harshbarger asked Mr. Bhargava to give examples of successful AI sandboxes and pilot programs. Mr. Bhargava cited Hippocratic AI, a startup aiding overburdened nurses, and recommended collaborative government-industry sandboxes to guide safe AI deployment, particularly in healthcare. Rep. Harshbarger followed up by asking about the Stargate Project and its role in strengthening America’s AI infrastructure. Mr. Bhargava emphasized that while infrastructure is key, the U.S. must also support its startup and university ecosystem, which he called the country’s “secret sauce” in AI leadership.

• Rep. Trahan requested unanimous consent to enter a civil rights coalition’s letter opposing the AI moratorium into the record and criticized the false dichotomy between innovation and regulation. Mr. Bhargava agreed that a balanced U.S. innovation environment requires immigration, research investment, and regulatory frameworks that are both protective and innovation-friendly. Rep. Trahan then asked what effective AI transparency requirements would look like. Mr. Bhargava said transparency must be meaningful and implementable, noting that overly burdensome processes in Europe have slowed feedback and stifled companies, and a better balance is needed in U.S. policy.

• Rep. Obernolte asked whether the Chamber of Commerce supported a federal moratorium on state-level AI regulation. Mr. Heather said yes, arguing that the U.S. should avoid both international and domestic regulatory patchworks and take a deliberate, federal approach.

  Rep. Obernolte defended the moratorium as a bipartisan, necessary step to prevent regulatory chaos, stating that the intent was to protect small innovators, not Big Tech, and that existing consumer laws remain enforceable under the bill.

• Rep. Soto asked what should be done to protect children from AI harms, referencing a tragic case in Orlando. Ms. Kak called for preventative red lines, transparency rules, and limits on business models that exploit behavioral data, warning that current frameworks are failing to protect young users.

  Rep. Soto concluded that Congress must find a middle ground that protects children while enabling AI-driven productivity, noting that legislative action—even if imperfect—must move forward.

• Chairman Guthrie asked what the impact of EU-style regulation would be on small startups and innovators. Mr. Bhargava responded that while European regulations aim to protect consumers, many rules are impractical, such as requiring error-free datasets, and often reflect a disconnect from real industry practices.
  Chairman Guthrie warned against models that require FDA-style approval for AI, noting that such delays could harm U.S. competitiveness, and asked about the risks of disclosure mandates in the EU AI Act. Mr. Heather said the EU AI Act’s broad disclosure requirements could expose U.S. intellectual property to competitors, including China, and would discourage investment due to the erosion of proprietary advantages.

  Chairman Guthrie then asked why state governors, including Colorado’s, have begun pushing back against AI laws. Mr. Thierer explained that Colorado's law prompted backlash from entrepreneurs, lacked clarity, and has since led to calls for a moratorium, while European businesses are spending enormous resources just to comply with AI regulations, which deters innovation.

• Rep. Mullin argued that regulation and innovation are not mutually exclusive and asked what strong, enforceable AI transparency requirements would look like. Ms. Kak replied that effective transparency should span the AI supply chain, require disclosures from both developers and deployers, and include infrastructure metrics like data center resource use, stressing that common-sense rules are being resisted by Big Tech.

• Rep. Bentz asked what the best investment strategies are to fund AI innovation without harming consumers and whether the current U.S. model is sustainable.
  Mr. Bhargava responded that AI automation in services is already producing value, similar to past investments in cloud computing, and justifies the current scale of investment. Mr. Thierer emphasized that smart regulation is essential to maintain private investment flows into the U.S., and warned that Europe lost capital and talent due to overregulation. Ms. Kak cautioned against inflated promises about AI breakthroughs like curing cancer, noting that such claims often lack evidence and coincide with cuts to actual research institutions like NIH. Mr. Heather concluded that the U.S. model of incentivized private investment remains the most effective, particularly compared to Europe’s limited capital and China’s state-driven approach.

• Rep. Eshoo criticized the 10-year moratorium on state AI regulation, warning that it would leave children and consumers unprotected, and asked what urgent protections Congress should adopt. Ms. Kak recommended bright-line rules to ban egregious AI harms like deepfakes, emphasized end-to-end accountability for developers and deployers, and endorsed mandatory impact assessments to verify system claims and performance. Rep. Fry framed AI as the next transformative infrastructure and asked Mr. Heather what mistakes the EU made with its AI regulations. Mr. Heather responded that Europe overregulated by failing to assess existing legal frameworks, overclassifying AI risks, and focusing excessively on Big Tech while ignoring B2B innovation from smaller firms.

• Rep. Fry asked if the U.S. risks creating a permission-based innovation culture. Mr. Heather said states appear to be heading in that direction and highlighted civil society concerns that the EU AI Act gives governments more AI power than the private sector. Rep. Fry then asked how regulatory delays affect innovation in fast-moving AI sectors. Mr. Bhargava warned that because AI models evolve every three to six months, excessive regulation—even short delays—can leave the U.S. at a technological disadvantage.

• Rep. Schrier asked about the usefulness of tools like model cards to disclose AI system limitations. Mr. Bhargava supported standardized tools and outlined four framework areas—data collection, training methods, model output testing, and downstream impact disclosure—calling for simple federal standards developed with industry input. Rep. Schrier asked what lawmakers should avoid in drafting transparency rules. Mr. Bhargava advised against writing rules in a vacuum or targeting Big Tech rhetorically, instead urging consultation with startups and market participants. Rep. Schrier then asked whether such disclosures should be voluntary, required for high-risk use, or mandated broadly. Mr. Bhargava said they should be mandatory but minimal, cautioning against burdensome audits that stall startups due to poor agency responsiveness. Rep. Schrier asked if NIST should lead voluntary AI standards. Mr. Thierer said NIST has already done valuable bipartisan work on AI risk management and Congress should build upon that.

• Rep. Clarke criticized the moratorium as reckless, citing the lack of a federal privacy law, and asked why federal privacy protections are critical to AI regulation. Ms. Kak emphasized that AI amplifies harmful surveillance practices, called for strong federal data minimization standards, and argued that privacy laws would help both consumers and competition by limiting Big Tech’s consolidation strategies.

• Rep. Kane asked how to “future-proof” federal AI policy. Mr. Bhargava recommended focusing on flexible national guidelines for transparency, data use, testing, and red-teaming, noting the need for policies that evolve with rapidly advancing technology. Rep. Kane then asked about the impact of state regulatory patchworks on innovation. Mr. Bhargava said they create significant burdens for startups without large legal teams and emphasized that a national framework is vital to U.S. competitiveness across multiple AI-enabled sectors.

  Rep. Kane asked how implementation of the EU AI Act is progressing. Mr. Heather said the EU faces major ambiguity and enforcement challenges, predicting that the complexity will overwhelm smaller businesses while even large firms will struggle with compliance.

  Rep. Kane asked Mr. Thierer to explain the rationale and precedent for a federal moratorium on state AI regulation. Mr. Thierer cited the 1998 Internet Tax Freedom Act as precedent and explained that a moratorium would create space for federal policymaking while preventing a chaotic patchwork of conflicting state laws.

• Rep. Veasey entered a Washington Post article on New Orleans’ secret use of facial recognition into the record and asked if, in light of a proposed 10-year moratorium on state AI laws, Congress should establish clear federal rules for AI in law enforcement. Ms. Kak agreed and emphasized that state efforts have addressed high-impact areas like criminal justice and education, adding that eliminating these protections without replacement would undo meaningful progress. Rep. Veasey then asked what Congress could do to ensure AI technologies are used ethically without infringing on civil liberties. Ms. Kak called for transparency, vetting, and accountability, citing the FTC's action against Rite Aid for misusing facial recognition as an example of necessary enforcement. Rep. Veasey followed up, asking how the U.S. can maintain leadership while protecting civil liberties. Ms. Kak said the U.S. should incentivize a “Race to the Top” by eliminating predatory actors and creating an AI sector built on trust and integrity.
  

• Vice Chairman Joyce raised concerns about the state patchwork of AI laws and asked how AI could improve patient care. Mr. Bhargava described General Catalyst’s investments in AI tools like medical scribes and workload reduction software, stressing the need for both innovation and responsible regulation. Vice Chairman Joyce asked if the current state-level patchwork puts patient lives at risk. Mr. Bhargava said the patchwork burdens startups and urged a federal framework to accelerate life-saving AI deployment.
  Vice Chairman Joyce asked how the EU treated existing sectoral laws before passing the AI Act. Mr. Heather said the EU didn’t identify gaps in laws for medical devices or pharma and warned that overregulating the technology itself could harm innovation and competitiveness.

• Rep. Evans shared his opposition to Colorado’s state AI law and asked Mr. Thierer to explain why such patchwork is harmful. Mr. Thierer said Governor Polis appeared to regret signing the law and noted that ambiguous definitions and complexity make it unworkable for small businesses.
  Rep. Evans asked about the fiscal impact of a patchwork approach. Mr. Heather cited the high cost of compliance per regulation, using a €330,000 estimate from the EU AI Act as an example, and warned that similar U.S. efforts would be expensive but lacked good current cost data. Rep. Evans asked whether national standards could generate more industry revenue. Mr. Heather said unleashing AI could yield trillions in economic growth, which would also increase tax revenues.
  Rep. Evans concluded by asking whether the proposed federal moratorium still allowed for child protection. Mr. Thierer affirmed that technology-neutral laws would still permit enforcement actions and stated that the moratorium's goal is to reduce legal barriers to responsible AI deployment.

ADD TO THE NIMITZ NETWORK

Know someone else who would enjoy our updates? Feel free to forward them this email and have them subscribe here.