The US Has No Real Plan to Regulate AI, and That Should Worry You
Tech CEOs and their own researchers are sounding the alarm on AI risks. The federal government's answer, so far, is to do nothing.
Key points
- Sam Altman, Dario Amodei and Elon Musk all publicly called for slower AI development over the same weekend, citing growing safety risks.
- At least one OpenAI researcher resigned publicly, accusing OpenAI and Anthropic of "gambling with our lives".
- The Trump administration and Republican leadership have pushed for AI self-regulation, meaning companies would set their own safety rules.
- Real-world cases of AI behaving unpredictably are mounting, yet no federal AI safety legislation is advancing.
It's a strange moment when the people building the most powerful technology on earth start begging someone to slow them down. Over the same weekend, three of the biggest names in AI, Sam Altman of OpenAI, Dario Amodei of Anthropic and Elon Musk of xAI, all said publicly that AI development needs to pump the brakes because the risks are getting serious. They're not outside critics. They're the CEOs building the thing.
As we reported on 13 September, Altman and Musk backed Amodei's warning that unchecked AI agents could seize control of the internet within a year. Their own staff are even louder about it. At least one researcher quit OpenAI in a public statement, accusing both OpenAI and Anthropic of "gambling with our lives." That's not a phrase you expect to see in a resignation letter from a tech employee.
So what is the government actually doing?
Not much. The Trump administration and Republican leadership favour what they call AI self-regulation, where companies write their own safety rules rather than following laws passed by Congress. The Guardian reported on the widening gap between the urgency of these warnings and the absence of any federal response.
Self-regulation sounds reasonable until you think about it for ten seconds. A company racing to beat competitors has every financial reason to skip the safety checks that slow it down. Asking that same company to police itself is a bit like asking a bakery to run its own health inspections: you already know how that ends.
What does this mean for ordinary people?
More AI is landing in your life every month. Doctors' offices use it to triage patients. Banks use it to approve or reject loans. If the software doing those jobs behaves in ways nobody fully understands, and there's no watchdog with teeth, the people who pay the price aren't the CEOs. They're the patients and the account holders.
When AI makes a decision that affects you, you have the right to ask a human to review it. In the US, the Equal Credit Opportunity Act already gives you the right to know why a credit decision was made. Ask for a human reviewer. Use that right.
Common questions
What is AI self-regulation and why is it controversial?
Self-regulation means AI companies create and enforce their own safety guidelines without binding laws. Companies face strong financial pressure to move quickly, which can push safety concerns aside when no external rules force otherwise.
Can an AI decision about me be overturned by a human?
In many cases, yes. In the US, laws like the Equal Credit Opportunity Act give you the right to ask why a credit decision was made, and requesting a human reviewer is a practical first step if an automated system affects you in a serious way.
My read: the warnings coming out of Silicon Valley aren't theatre. When rival CEOs who rarely agree on anything concede their own industry needs reining in, that's worth taking seriously. Washington's silence is the real story here, and it won't stay comfortable for long.
