The "One Big Beautiful" Bill Threatens a Decade of AI Accountability

The House Republicans' narrow passage of H.R. 1—dubbed the "One Big, Beautiful" bill—contains a provision that should alarm everyone working toward the aims and ambitions of developing an ecology of responsible technology: a 10-year moratorium on state enforcement of AI regulations. 

By Deb Donig, Siegel Research Fellow for All Tech Is Human

For the ATIH community, this represents a potentially devastating rollback of the accountability frameworks we've spent years building, so it is worth knowing what this means for our work–and potentially for your career. Below, I outline the legislation and its potential impact.

The moratorium, as proposed, would effectively freeze the current patchwork of state AI laws that address critical harms our community has long identified: discriminatory hiring algorithms, deepfake abuse, biased decision-making systems, and inadequate transparency in automated systems. 

States like California, New York, and Illinois have been laboratories for AI governance, developing targeted solutions for algorithmic bias in employment, housing, and criminal justice—exactly the kind of sectoral approach that responsible tech advocates have championed as more effective than broad, toothless principles. 

As laboratories, the legislation passed in this state functions in two critical ways: first, by passing laws, favored by representatives of the people, through democratic processes and by creating protections for civilians in the present; second, by creating experimental paths forward, where ideation, impact, and efficacy can be assessed, provided trials, and – if successful – be elevated to national policy. Crucially, as laboratories, the institution of these policies on the state level provide populations an example fo what legislation could or should look like, thereby allowing populations an understanding of what could be, and what the vision of responsible technological legislation could look like. Without that vision, it is hard for populations to understand what they may gain or lose through the democratic processes of legislation, and what a vision of responsible technological governance could look like, a key component of building popular support for future legislative approaches.

The legislation comes at a moment when our field is gaining momentum despite earlier setbacks. The Biden administration's AI Executive Order, state-level algorithmic accountability laws, and growing corporate recognition of AI risks had created new opportunities for responsible tech professionals. A 10-year freeze would halt this progress precisely when AI systems are becoming more powerful and pervasive.

Although the concerns about the needs for proper legislation tend to belong to the broader Democratic wing of Congress, concern over this legislation is, on some level, shared across party affiliation, as bipartisan. Senator Marsha Blackburn's concerns about Tennessee's ELVIS Act—protecting artists from AI-generated deepfakes—illustrate how a blanket moratorium could eliminate targeted protections that enjoy bipartisan support. 

While the bill faces uncertain prospects in the Senate, its passage in the House signals the tech industry's continued resistance to accountability measures. The administration’s ties, and its obligations, to tech leaders who have vocally opposed legislation limiting the development and deployment of AI means that legislation favorable to financial and other interests of tech companies has greater likelihood of garnering favor. 

For our community specifically, this isn't just about regulatory policy—although that is what many of us work tirelessly in pursuit of. But it’s also about the infrastructure supporting our professional community. Many responsible tech roles exist precisely because of compliance requirements from state laws like California's AB 2273 (protecting children online) or New York City's automated employment decision tools law. A federal moratorium could eliminate the regulatory foundation that justifies these positions, potentially triggering another wave of layoffs in ethics and safety teams reminiscent of 2022's decimation.

For those of us in the US, this is a moment to pay attention and to engage. The fight isn't over, but it's a reminder that building a responsible tech workforce means more than thinking narrowly about tech products, or even the tech industry. It also means participating in building political power to defend the regulatory foundations that make our work possible. The next few months will test whether the responsible technology movement can mobilize effectively beyond industry circles into broader democratic participation.

For those in the ATIH community working in responsible tech outside the United States, this legislation matters more than it might initially appear. American tech companies operate globally, and their compliance strategies—or lack thereof—inevitably shape practices worldwide. When US companies can avoid accountability measures for a decade, that affects the products and services deployed in your countries too. 

We've seen this pattern before: when Facebook faced European GDPR requirements but lighter US oversight, the company often implemented stronger protections in Europe while maintaining more invasive practices elsewhere. A 10-year US moratorium could create similar dynamics on steroids, where American companies become even more resistant to accountability measures internationally, arguing that if the US—their home market—doesn't require certain protections, why should other jurisdictions? Moreover, many of you work for US-headquartered companies or their subsidiaries, meaning your responsible tech roles could be directly affected by the elimination of compliance requirements that justify these positions. This legislation may be American governance in scope, but it is global in impact, in consequence, and in its meaning for our community, which is intentionally and importantly international. In this moment, we may see whether the global responsible tech community, which has developed international cooperation and idea pipelines across borders can maintain the momentum we've built across multiple jurisdictions, or whether a decade of US companies will, ultimately, actively undermine international accountability efforts while pointing to this moratorium as justification.

A decade without state-level AI accountability could entrench harmful systems at unprecedented scale, making our eventual work of reform exponentially more difficult. It will, almost certainly, transform the global landscape as well. For the US, the impact of AI on democratic processes will be part of that consequence; it is of consequence to understand and engage in the democratic processes around AI. For the international community, our ability to cooperate, and share contexts and premises of development across borders may have just gotten more complicated–and more important.