Dear Representative:
On behalf of the 63 affiliates of the American Federation of Labor and Congress of Industrial Organizations (AFL-CIO), representing 15 million working people across our economy, I write to draw your attention to a particularly reckless provision included in the “One Big Beautiful Bill Act” committee print that would bar states from enforcing regulations related to artificial intelligence (AI) or automated decision systems. This federal preemption provision would override state and local authority, stripping our communities of the ability to safeguard workers and the general public against harmful and unregulated AI use. It should be stripped from any final bill. It is one of many reasons to oppose the current bill.
In the states, lawmakers have come together to protect civil liberties, advance worker protections and prevent discrimination in the artificial intelligence context. All 50 states have introduced legislation or regulations related to AI, algorithms, and machine learning, and around 40 have passed legislation. Many of these legislative efforts have been bipartisan. The bill’s proposed moratorium would nullify enforcement of these and any future state laws across the country, unless they fall into very narrow exceptions. States are the laboratories of democracy, and banning existing and future policymaking will shut down those laboratories and leave everyone vulnerable to wide-ranging abuses stemming from irresponsible AI.
The states’ policymaking efforts are critical. While AI has the potential to boost productivity, it also has the potential to cause great harm to workers if left unregulated. The examples are manifold:
- Transportation workers are seeing an onslaught of AI-enabled autonomy that opens the door to eliminating the human in charge despite the safety threats to the traveling public.
- Health care workers, such as nurses, find their schedules and patient assignments determined by secret algorithms that put speed over quality of patient care and worker well-being.
- Workers in creative industries, such as journalists, behind-the-scenes film and television crew, writers and performers, see their work, including their likeness, being stolen by generative AI. In recent years, AI has come to directly threaten the livelihoods of creative workers and the integrity of their professions.
- Schoolteachers face a flood of AI technologies targeting education and even attempts by some to substitute their skill and judgment with AI technology.
- Federal workers have faced mass layoffs, with brazen attempts to replace their deep expertise with shoddy AI systems.
- Warehouse workers face technologies that create unsafe working conditions and are monitored in bathrooms at work and private spaces outside of work.
- Discriminatory hiring algorithms and surveillance systems can unfairly put entire groups of people – based on race, gender, or other protected characteristics – at a permanent disadvantage. Unregulated AI can destroy equal opportunity for all.
Artificial intelligence systems are evolving at a rapid pace. It is difficult to predict what they will be capable of or what new harms will befall people from them next year, let alone over the next ten years. The moratorium suggests, however, that there is nothing to worry about, and we should put the interests of Big Tech over the concerns of ordinary people. The preemption is so broad that it even threatens state safety regulations for the kinds of automated systems employed in public transit. The moratorium contains only two exceptions. One is so narrow that it would not save from preemption any actual guardrails against unsafe or irresponsible AI. The other exception, added after the initial markup of this provision, is a poor attempt at addressing the deadly outcomes promised by a blanket moratorium. It exempts from federal preemption state regulations that carry a criminal violation. So, under the bill, if a state wants to address any problem at all with AI use, it must criminalize the matter, locking ordinary people out of having rights they might enforce and simply relying on criminal prosecution each and every time. In lieu of encouraging nuance, the current bill now expects AI innovation to thrive under a new system where the state’s only tool to address an issue is a sledgehammer. This is no way for industry to innovate or for policymakers to effectively and responsibly protect their citizens.
Federal preemption would endanger progress in the states and silence local voices responding to real-world harms from worker surveillance and algorithmic firing, to automated price and wage manipulation that undermines economic fairness, to deepfakes and disinformation that undermine safety and public trust. With Congress failing to enact meaningful AI regulation, states remain our last line of defense. I urge you to oppose this dangerous provision and allow states to do their jobs.
Sincerely,
Jody Calemine
Director, Government Affairs