Republicans in the U.S. Congress have tried to ban states from introducing or enforcing laws to create a guardrail of 10-year-old artificial intelligence or automated decision-making systems.

Now, a provision in the proposed budget bill would prohibit any state or local government agency from taking “any law or regulation that regulates AI models, AI systems or automated decision-making systems” before the House, unless the purpose of the law is to “remove or promote the deployment of these systems or promote legal barriers.”

The rule is the last minute that House Republicans joined the bill before the two nights marked on Tuesday. The House Energy and Commerce Committee voted to advance the reconciliation plan Wednesday morning.

The bill broadly defines AI systems and models, from facial recognition systems to generating AI qualifications. The proposed law also applies to systems that use algorithms or AI to make decisions, including hiring, housing, and whether to qualify for the public interest.

Many of these automated decision-making systems have been under fire recently. The proposal for the relaxation plan follows a lawsuit filed by several state attorneys generals against property management software RealPage, which alleges merged with landlords to raise rents based on the company's algorithm recommendations. Another company, the safer company, recently settled a class action lawsuit filed by Black and Hispanic renters who said they were denied apartments based on the opacity score the company gave them.

Some states have signed laws that will attempt to establish safeguards around these systems. For example, New York passed a law that requires automatic recruitment systems to conduct bias assessments. California has passed several laws that regulate automated decision-making, including one that requires healthcare providers to notify patients when sending communications using Generative AI. If the settlement bill passes, these laws may not be enforced.

"The bill is a daunting and reckless attempt to protect some of the world's largest and most powerful companies - from large-scale technology monopolies to reality pages, any responsibility of UnitedHealth Group and others," said Lee Hepner, senior legal counsel for the U.S. Economic Freedom Project.

The new language is in line with the Trump administration's actions and aims to remove any obstacles to the development of AI. After taking office, Donald Trump immediately revoked a Biden administration execution order that created a security guardrail for the deployment and development of AI. Silicon Valley has long believed that any regulations stifle innovation, and several prominent members of the tech industry joined the U.S. presidential campaign, leading the administration to echo the same point.

“State legislators across the country are stepping up real hazard solutions – a bill that is a pre-strike strike that can close those strikes before they get more foundations,” Hepner said.