The key question is, what’s the future of “‘morally binding’ self-policing” if it is left to AI moguls?

From The Hill we learn,

In a rare White House North Lawn appearance, the president and House speaker were joined by numerous tech executives — including Nvidia CEO Jensen Huang, OpenAI president Greg Brockman, Anthropic CEO Dario Amodei, Google CEO Sundar Pichai, Meta CEO Mark Zuckerberg and xAI CEO Elon Musk.

“The basic idea is that we want to give the American people and our customers confidence that the technology works in the way that we intend,” Zuckerberg said.

As part of the accord, the companies are committing to building “robust” internal controls and detection capabilities, the Meta CEO said. This is coupled with “multiple layers of auditing.”

Miranda Nazzaro and Julia Shapero, “AI firms agree to ‘morally binding’ self-policing at White House ,” September 29, 2026

And if the moguls don’t self-police well?

Reactions were cautious, if not suspicious:

From EuroNews, “The accord however does not introduce a set of legislation that introduces robust legal liability and regulation to guide the technology, which some in the industry have blown the whistle on, saying there is a chance it could outsmart humans and cause their extinction within a decade.”

That question about “robust legal liability” could prove important. Ignoring the sci-fi trimmings, one big question is, who assumes liability in the event of a major computer security disaster?

From The Guardian, “None of the controls listed in the agreement involve government regulators or include a commitment to publicly detail the findings of independent evaluations. The agreement also appears to allow companies to pick their own evaluators, appoint their own oversight boards and decide whether to publish their results.”

Here, the question is, would they have agreed to anything they did not completely control? And then what exactly does the agreement rule out?

AI analyst Gary Marcus offers the text of the Accord and comments,

Reading the Accord between the lines

We agree not be regulated

We agree not to give the public a voice

Trust us

One thing about the accord: It is certain to be tested by events.