00:00U.S. President Donald Trump and the leaders of some of the largest artificial intelligent companies have signed a set
00:06of voluntary safety standards, promising to self-police the technology.
00:12Trump announced the move after hosting top AI executives at the White House, as his administration continues to resist calls
00:19for tighter guardrails on AI technology.
00:21Well, I think I'm seeing tremendous self-policing, and they understand that they have to self-police. And it's very
00:27hard for somebody to come outside, get into these models. These models are very complex. And again, I believe they're
00:34going to be used for the good, and when they're not, we're going to be able to nab them. But
00:37there's going to be a tremendous self-policing aspect.
00:41Those present included leaders from OpenAI and Tropic, NVIDIA, SpaceX, Meta, and Google. Under the agreement, the tech companies committed
00:50to ensuring the AI models are acting as intended, including internal evaluations, audits by an external firm, and reviews by
00:58each company's board of directors.
01:01Trump said the agreement is morally binding and not legally enforceable, but leaves the door open for the measures to
01:08later be codified into law or regulations.
01:11The move followed months of growing concerns from researchers, politicians, and industry figures that ever more powerful AI technology could
01:19pose safety risks.
01:20Both OpenAI and Tropic have reported instances of their AI models going rogue and gaining unauthorized access to other companies'
01:28systems or being used for dangerous purposes.
Comments