Leaders pledge to contain potentially 'catastrophic' AI risks at U.K. summit Government officials and industry leaders convened an international summit in the United Kingdom to examine possible regulation of artificial intelligence.

Leaders pledge to contain potentially 'catastrophic' AI risks at U.K. summit

  • Download
  • <iframe src="https://www.npr.org/player/embed/1210000802/1210000803" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript

MARY LOUISE KELLY, HOST:

Delegates from 28 countries, including the U.S. and China, gathered today at a bucolic English country estate. They gathered to sound an alarm about AI - artificial intelligence. They pledged to work together to contain what they call the potentially catastrophic risks posed by this technology. Well, NPR's Lauren Frayer has been covering this two-day summit of politicians and tech CEOs. Hey there, Lauren.

LAUREN FRAYER, BYLINE: Hi.

KELLY: I want to start with a very basic question. When we say they're gathered; they're talking AI - what do they mean? What has them so worried that they've all come together?

FRAYER: Yeah, so there's a lot of, like, talk about existential threats. And the U.K. prime minister, Rishi Sunak, has actually said he's literally worried about humanity losing control of computers. Delegates are talking about the role of AI in bioweapons, cyberattacks, making the global financial system go haywire, threats to democracy. But AI is also stuff that, like, you might be familiar with - language models, things like ChatGPT, Google Translate, and also tools that, like, your bank may be using to evaluate loan applications, read your medical records.

Vice President Kamala Harris is in town for this summit, and she gave a speech today at the U.S. Embassy in London where she gave some real-world examples.

(SOUNDBITE OF ARCHIVED RECORDING)

VICE PRESIDENT KAMALA HARRIS: Consider, for example, when a senior is kicked off his health care plan because of a faulty AI algorithm. Is that not existential for him?

FRAYER: So she talked about people being victimized by deepfakes - like, fake explicit photographs - or people even thrown in prison because of biased AI facial recognition software.

KELLY: OK.

FRAYER: And so she talked about the need to regulate all of that stuff without stifling innovation.

KELLY: Yeah, that's the challenge. OK, so we've got the American vice president there, the British prime minister - Prime Minister Sunak - everybody talking about it. I guess the question that's always hanging over this is what can they actually do? What are they going to do?

FRAYER: Yeah, so they can issue a joint declaration, which they did today, but, you know, there is no global regulatory body. There's nothing binding here. This summit is a success in terms of just bringing these people together. And it goes beyond politicians. I consulted an AI expert, Nina Schick, and she says a major role is going to have to be played by U.S. tech companies, too.

NINA SCHICK: Primarily because the knowledge in terms of what is actually happening - how are these systems built? What's under the hood? Who can understand them? - is not going to come from government alone. Industry needs to be a part of this conversation.

FRAYER: And so to that end, Elon Musk is participating. Representatives of Google, Amazon, Meta, Microsoft are all attending this summit.

KELLY: Well, and that prompts me to ask you about the venue. I mentioned at the top they were all meeting in this somewhat unlikely setting - a bucolic English country estate. This is not just any English country estate, though. Explain.

FRAYER: It's gorgeous. It's called Bletchley Park. It's this sprawling brick-and-stone manor house, and it was once the top-secret home of World War II codebreakers, where some of the sort of best and brightest mathematicians, including one named Alan Turing, cracked the Nazi's Enigma code and helped win the war. The 2014 movie "Imitation Game..."

KELLY: Yeah.

FRAYER: It's there. And so Bletchley had a role in the birthplace of some of the earliest programmable computers. And the questions that Alan Turing asked at Bletchley literally 85 years ago - can computers think? Can they imitate humans? - are being talked about again in that very same venue.

KELLY: Fascinating. NPR's Lauren Frayer...

FRAYER: It is.

KELLY: ...In London. Thank you.

FRAYER: Thanks so much.

Copyright © 2023 NPR. All rights reserved. Visit our website terms of use and permissions pages at www.npr.org for further information.

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.