SACHA PFEIFFER, HOST:
At the White House today, President Trump hosted top artificial intelligence leaders amid an ongoing debate over the safety of this technology that is advancing at a head-spinning pace. For weeks, AI executives have been issuing dire warnings about the catastrophic risks of their products and stories of AI agents going rogue. Now, the Trump administration wants to change that conversation. President Trump brushed off those warnings at the launch of a new AI-powered government website. He also promised that the technology will usher in a new age of prosperity.
(SOUNDBITE OF ARCHIVED RECORDING)
PRESIDENT DONALD TRUMP: We do not surrender to fears about what technologies will do to us. We decide that those technologies will do for us some very, very incredible things.
PFEIFFER: NPR's Huo Jingnan covers AI, and she's here to walk us through the latest. Hi.
HUO JINGNAN, BYLINE: Hello.
PFEIFFER: This meeting today at the White House with AI leaders, what do we know about what happened at it?
HUO: So we know that the guests include leadership of most major tech companies in this country. That includes Google, Meta, OpenAI and Anthropic. And Trump described, quote, "a lot of commonality in the room." But actually, the industry is divided, that (ph) the executives here have a wide range of views on AI safety. For example, OpenAI and Anthropic have been more outspoken about safety concerns. They have called for a coordinated slowdown in the development of the most advanced AI systems. But others, including Nvidia and Meta, have argued against excessive regulation and did not support a coordinated industry slowdown. So despite what the president says, there is not a common vision on what should be done about the potential risks of AI, even though the president himself says self-regulation was the most appropriate approach.
PFEIFFER: This meeting today was not the only meeting the president had related to AI. Give us a sense of what other AI meetings have been on his schedule.
HUO: So Trump has had dinner with Anthropic CEO Dario Amodei on Sunday, which is interesting because Anthropic has disagreed with the White House about a range of things, including regulation and how much AI companies can restrict the use of their products. And this morning, Trump and many of these tech executives at the luncheon were at the public launch of a new AI-powered website called america.gov that's supposed to be a more accessible portal to all kinds of government services. They are promising that you'll be able to use it for anything from getting a passport to booking a camp site. And Trump also says he's renaming AI to, quote, "super intelligence."
PFEIFFER: Super intelligence - how is he defining that?
HUO: Well, President Trump says artificial intelligence has a branding problem, so he said that the word artificial isn't helpful and that it sounds fake. So he's dropping the word, and he's, you know, already used a new term, super intelligence, in his latest executive order today about the launch of the AI-powered website. You just don't see the word AI on there, just super intelligence. You know, overall, Trump seems to be trying to convince the public that AI is good for them, even as people are increasingly worried about losing jobs to AI, the impact of data centers and the safety risks that tech poses.
PFEIFFER: So two things are happening simultaneously. The White House is powering ahead on AI, but yet we hear this drumbeat of alarming safety incidents. Recap some of the things that people are concerned about.
HUO: Right. So this week, OpenAI actually canceled the release of its most advanced model, and it also paused training last week after its models gained improper access to the internet. And this comes after we learned last week about even more incidents in the past few months of OpenAI systems doing things autonomously that people wouldn't expect them to. And that includes hacking into an Australian government website and using other people's credentials to access publicly accessible U.S. census data.
But look, part of the resistance here from Trump and some AI executives is that they want to make sure that the U.S. stays ahead of China when it comes to AI development. And many in the AI safety field are hoping that the two countries will move towards a deal to collaborate on AI safety. There have been some talks between the two countries about AI in the past couple of weeks, but there has been no agreement so far. So bottom line is that there's a lot of calls from the AI safety field for U.S. government to regulate the industry and set up and enforce some standards. But as today showed, there's currently little sign that this White House is interested in that.
PFEIFFER: That is NPR's Huo Jingnan. Thanks.
HUO: Thanks so much.
PFEIFFER: And we want to note that the AI company Anthropic is a financial supporter of NPR.
(SOUNDBITE OF MUSIC) Transcript provided by NPR, Copyright NPR.
NPR transcripts are created on a rush deadline by an NPR contractor. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.
