AI bosses sign self-regulation accord

Charlotte Gallagher (0:00)

This is The Global News Podcast from the BBC World Service.

I'm Charlotte Gallagher, and in the early hours of Wednesday, the 30th of September, these are our main stories. As warnings continue over the risks artificial intelligence may pose to humanity, Donald Trump says US-based AI bosses have signed an accord on self-regulation that he's called morally binding. South Africa's president, Cyril Ramaphosa, addresses the nation amid growing alarm about violence against women. And the English Premier League confirms Manchester City have been found guilty of serious breaches of financial rules. Also in this podcast, the UK's Prime Minister, Andy Burnham, sets out an ambitious agenda in a major speech to his party.

Andy Burnham (0:49)

Deregulation, privatisation, austerity, Brexit. And what has it given us?

A country run for vested interest, not in the public interest.

Charlotte Gallagher (1:00)

And?

Julian Gascoigne (1:01)

This is one of the last great masterpieces by him, remaining in private hands. And it's a view of Rome.

Charlotte Gallagher (1:08)

A rare painting by the renowned artist Turner returns to the Eternal City.

The Chinese AI company Moonshot is conducting an internal review after researchers said he persuaded its technology to discuss making chemical and biological weapons and how to kill people. It was only a few weeks ago that a US AI expert said the technology was advancing so quickly. He believed there was more than a 10% chance it could kill all humans within the next decade. But now President Trump has said tech companies developing AI can regulate themselves. He was speaking after talks with industry leaders at the White House. Mr Trump said the tech leaders had signed what he called a morally binding accord on AI regulation.

Donald Trump (2:01)

Well, I think I'm seeing tremendous self-policing.

These models are very complex. And again, I believe they're going to be used for the good. And when they're not, we're going to be able to nab them. We're also thinking about forming a committee of sorts where we put maybe ten people on that committee. It could be from that group so the committee can watch over the whole enterprise. We're looking at this for good. It's also maybe the biggest industry. They say bigger than the industrial revolution.

Charlotte Gallagher (2:28)

Among those attending the White House meeting were the Anthropic boss, Dario Amadei, Open AI's Greg Brockman, Nvidia's Jensen Huang and even Elon Musk, apparently back on good terms with Donald Trump after that very public fallout. Here's Metta's founder, Mark Zuckerberg.

Mark Zuckerberg (2:47)

I think what we can all agree on is that the American public and our customers want to know that what we're building is working in the way that we all intend. So we drafted a set of principles and commitments around building robust internal controls and detecting if there are any issues with the technology, coupled with multiple layers of auditing and controls and agreeing that we're going to have all of our boards of directors independently review the reports that come from the auditors. It was sort of a historic conversation. I don't think anything like this has come before.

And I think this is a meaningful step forward.

Charlotte Gallagher (3:22)

So what exactly is in the accord? Burt Deboosman is our White House correspondent.

Burt Deboosman (3:28)

This accord is essentially a set of responsibilities, not quite regulations, but things that the industry internally and collectively needs to address as far as some of these safety concerns that have come with AI. Now, President Trump was asked directly whether this would be in any way a legally binding agreement, and he said, no, it's a morally binding agreement. So I think to many of the people that are concerned about the future of AI and what it could mean for society, that'll raise some eyebrows because at the end of the day, there's really nothing other than the word of these tech CEOs that were here at the White House that enforces that mechanism. It's really just a collective agreement to keep an eye on those responsibilities in the future. President Trump very much does not want any sort of top down government regulation on artificial intelligence because he thinks that'll put the US at a disadvantage with regards to China and stifle what he thinks could be a huge engine of economic growth in the United States in the future.

Charlotte Gallagher (4:25)

Some people are going to be really cynical about this. Given that these are business people, they want to make money.

And you think of the number of court cases that are happening at the moment involving tech companies and then allegedly failing in their responsibilities.

Burt Deboosman (4:41)

And I think that's one of the reason President Trump and these tech leaders had this event to show that they're all on the same page with regards to that. Now today comes as polls show that a majority of Americans are concerned about the future of AI and as a related issue, AI data centers that are being set up in communities across the country. And President Trump really wanted to address those concerns head on to make the case that in the White House's view, this is something that could benefit Americans by creating jobs and high paying jobs in the future. And also that these companies would reinvest in the communities that they're building data centers in for the AI infrastructure. I think that's really a politically important point for President Trump to downplay the concerns about what this means for the future and try assure Americans that these for-profit companies ultimately will do the right thing and keep these guardrails in place.