Requires developers of advanced AI systems to submit them to a Department of Energy-run evaluation program testing for loss-of-control and weaponization risks before deployment, with penalties for noncompliance, and tasks DOE with generating data to inform a future federal oversight framework.
Vehicles carry asks; asks outlive them
Sorted by what filers wanted done to the rules
“Number one, I would form a new agency that licenses any effort above a certain scale of capabilities and can take that license away and ensure compliance with safety standards. Number two, I would create a set of safety standards focused on what you said in your third hypothesis as the dangerous capability evaluations.”
Speaker: Sam Altman, CEO, OpenAI. Third-party verbatim transcript.
“just come back to the question of, of whether we need an independent agency. I mean, I think we don't want to slow down regulation to address real risks right now. Right? So we have existing regulatory authorities in place who have been clear that they have the ability to regulate in their respective domains.”
Speaker: Christina Montgomery, Chief Privacy & Trust Officer, IBM. Third-party verbatim transcript.
“I think that would be disastrous. I to give a more specific answer to your previous question, which I think touches on why it would be so bad, there are three key inputs to these AI systems.”
Speaker: Sam Altman, CEO, OpenAI. Third-party verbatim transcript; transcript's 'I to give' artifact preserved.
“Congress has spent a lot of time over the last year debating whether to do away with regulations for the AI industry – this bill is a welcome show of bipartisan support for creating rules of the road to protect the public.”
“There is no doubt future administrations will attempt to do the same when it comes to AI. Unless further guardrails against this kind of graft are put in place or Congress finally acts to deliver a national AI framework, we are concerned the collaborative framework President Trump has put forward will not remain voluntary.”
Qualified support of EO 14409's voluntary approach; coded oppose relative to a mandatory review requirement.
“Voluntary frameworks are not enough, however. We need a mandatory government pre-deployment review process for the most powerful AI systems, allowing the government to block the release of systems that pose an unacceptable national security risk.”