Trump's 'Big Beautiful Bill' would create a regulation-free AI hellscape, AGs warn

[justaprogressive]

State attorneys general and activists are sounding the alarm over a provision of President Trump's budget proposal, which passed out of committee over the weekend and is headed to the House for a potential vote that would strip states of the ability to regulate AI.

The "One Big Beautiful Bill Act," Trump's proposed budget reconciliation bill for FY 2025, has spurred 40 state AGs, as well as more than 140 separate organizations, to write congressional leaders urging them to reconsider a particular passage about AI regulation.

Buried 291 pages into the 1,116-page bill as passed out of committee, Section 43201(c) describes a 10-year moratorium on state-level enforcement of "any law or regulation regulating artificial intelligence models, artificial intelligence systems, or automated decision systems."

Exceptions to the rule are few, and only allow for state-level rules that "remove legal impediments" to operating AI systems and streamlining their adoption. State laws that impose any substantive restrictions or requirements on AI models will be considered unenforceable under the decade-long moratorium.

https://www.theregister.com/2025/05/20/trump_bill_regulation_free_ai/

[mucholderthandirt]

1. Already happening. We creatives have no recourse for our stuff being used to train "AI". It's considered fair use now.

Thank you, justice system. There's an oxymoron for you.

We're all fucked, the country is fucked, and it will be fucked for generations.

[highplainsdem]

2. No, it isn't considered fair use - except by AI companies and their fans. But Trump got rid of the

US Copyright Office head who issued a report a few weeks back that the AI bros weren't happy with.

There's currently incredible lobbying pressure on the UK's Labour Party government to give the AI companies what they want, and if that lobbying succeeds despite creatives fighting back (mostly with Tory support, which says terrible things about Labour), that precedent will be cited here.

Of course AI companies have ALREADY stolen all the intellectual property they could, which is why they're fighting multiple lawsuits, and especially fighting having to be transparent about what they stole.

The only truly fair solution would be to destroy all current illegally trained AI models, and start over, training new models on only what's in the public domain and whatever else AI companies have arranged to use legally, by licensing or maybe paying for work on demand so they own the copyright.

As you can imagine, they have no intention of doing this. Some AI bros, including Musk, have even suggested getting rid of intellectual property laws.

Which makes it even more important for anyone with any ethical standards at all, who's aware of the IP theft, to refuse to use those AI tools and cooperate with that theft unless absolutely forced to by their work or school. It's also important for people NOT to circulate AI slop.

And it's critically important to remind people that these are fundamentally unethical tools that are causing a great deal of harm - dumbing down and deskilling users, making a shambles of education, hurting workers, and harming the information ecosystem and natural environment.