Disclosure up front: this post was drafted with Claude, made by Anthropic. Anthropic's CEO is one of the signers of the accord below. We've kept to AP's reporting and included critics in their own words.
Imagine if the companies building the most powerful technology on Earth agreed on a set of safety rules, then agreed to check themselves against those rules.
That's roughly what happened at the White House on September 29.
Who Signed
According to the Associated Press, the accord was signed by President Trump, Anthropic CEO Dario Amodei, Google CEO Sundar Pichai, Meta CEO Mark Zuckerberg, OpenAI President Greg Brockman, Nvidia CEO Jensen Huang, and Elon Musk, founder of xAI.
What They Promised
The accord, which Trump posted on Truth Social, focuses on voluntary steps. Each company will:
Implement "robust internal controls."
Partner with an "independent external auditor" to assess whether those controls are working.
Create a committee on its board of directors to evaluate reports from internal and external auditors.
And one sentence looks ahead:
"Over time, it may make sense to codify these steps into laws and regulations."
Trump said the accord would be "morally binding," and that about 10 people would be named to a committee to "watch over the whole enterprise." AP notes some of the steps are things these companies already do in some form.
What the Signers Said
Amodei, who AP describes as one of the most outspoken voices calling for a slowdown in AI development, said:
"The technology has very real risks. And, you know, the mechanism, how we address those risks is still under discussion."
Separately, OpenAI's Sam Altman told CNBC the company is "pacing our progress, which includes sometimes not training a model."
What Critics Said
Robin Jia, a USC computer science professor, is skeptical of putting "so much faith in self-policing," pointing to recent hacking disclosures: "As the models get better, the consequences of such things are only going to get greater."
Alex Pascal of Harvard's Berkman Klein Center said: "The only way to reduce the risks and harms of AI to Americans is through robust legal liability, regulation and fundamentally changing the race dynamics."
Shri Narayanan, a USC engineering professor, saw it differently: the accord balances the latitude needed for innovation, including innovation in safety, with regulation.
The Bigger Picture
Earlier that day, Trump said the government wouldn't support calls for AI "guardrails": "I will never stifle the growth of a technology that will be bigger than the industrial revolution."
Our view: for now, the main U.S. approach to frontier AI safety is self-governance with outside auditors. That's more than nothing, and less than law.
Why It Matters for Families
This is a great real-world example to explore with a teen: who should make the rules for a powerful new technology?
Three questions to talk through:
Would you let students grade their own exams if an outside teacher spot-checked them? That's close to how this accord works. What would make it fair?
Why might companies prefer voluntary rules over laws? And why might critics prefer laws?
What should happen if a company breaks a "morally binding" promise?
There's no single right answer, and that's the point. Teens will live with whatever rules get written in the next few years.
The Part Worth Remembering
The accord's own text leaves the door open to laws "over time." Until then, the companies building the most powerful AI are checking their own work.
Source: Associated Press, Christopher Rugaber, Will Weissert and Meg Kinnard, "Trump says top tech firms have signed accord to 'self-police' AI development," September 29, 2026. https://apnews.com/article/trump-ai-anthropic-musk-595796511f110fc006cca0d01329733e
Quotations verbatim from AP. The interpretation and family guidance are ours.
Disclosure: drafted with Claude, made by Anthropic, whose CEO signed this accord.
Want practical AI guidance for parents and educators every week? Subscribe: https://www.aibyage.com/?modal=signup&utm_source=beehiiv&utm_medium=newsletter