The White House and big tech companies release commitments on managing AI
Presidents and CEOs from leading tech companies that are creating artificial intelligence have agreed to several commitments on sharing, testing and developing new AI technology, the White House says.
Seven companies — Google, Microsoft, Meta, Amazon, OpenAI, Anthropic and Inflection — will meet at the White House Friday to announce the voluntary agreements. President Biden will also be making remarks.
"U.S. companies lead the world in innovation, and they have a responsibility to do that and continue to do that, but they have an equal responsibility to ensure that their products are safe, secure and trustworthy," White House chief of staff Jeff Zients told NPR in an interview.
But there isn't an exact outline for how the White House can hold the companies accountable if they don't meet the commitments.
"We will use every lever that we have in the federal government to enforce these commitments and standards. At the same time, we do need legislation," Zients said.
White House officials say they're working closely with Congress as they develop AI legislation that would regulate the technology, as well as working on executive actions that will be announced in the coming weeks.
What are in the commitments?
The commitments from tech companies are about information sharing, testing and transparency with both the government and the public. But there aren't many details offered in Friday's announcement.
For example, there's a commitment to develop mechanisms so that users will know when content is generated by artificial intelligence, through a watermark. Companies also said they would make a point to avoid bias and discrimination, and protect privacy.
Companies have also committed to having their AI systems tested through a third party before being released. One example of that will take place at the DEF-CON hacking convention in Las Vegas next month. Some of the companies, including Google and OpenAI, will have their AI systems tested there, at the encouragement of the White House. Beyond that, there isn't a clear outline of who the third-party checks would be, and how they are selected.
But the White House says these agreements are just a first step.
"The commitments the companies are making are a good start, but it's just a start," Zients said. "The key here is implementation and execution in order for these companies to perform and earn the public's trust."
Critics say big tech companies should not be the center of the conversation
Some AI experts have expressed concern about big tech companies being at the forefront of the conversation on regulating AI.
The White House has also been holding listening sessions with civil rights leaders and union leaders on how AI impacts their work.
But Ifeoma Ajunwa, a law professor at Emory who studies the intersection of technology and work, said she's found it disappointing that those who have a financial stake in AI development have been at the forefront of the White House's announcements on AI.
"We also want to ensure that we are including other voices that don't have a profit motive," she said. "We should definitely invite corporate leaders and tech titans to be part of this conversation, but they should not be leading the conversation."
There are also concerns that centering bigger, more established companies in the new agreements could give those businesses a leg-up, while stifling smaller companies that are just starting out.
"The bigger established firms can kind of game it to benefit them, and the newcomers don't have a say," said Victor Menaldo, a political economy professor at the University of Washington. "Big companies love to do these kind of things because they're already established, so they're like, 'Oh, the rules of the road are going to benefit us.' "
On a call with reporters Thursday, White House officials did not confirm whether new companies would be joining onto the agreements, in addition to the seven that have signed on.
"We expect that other companies will see how they also have an obligation to live up to the standards of safety security and trust. And they may choose — and we welcome them choosing — joining these commitments," a White House official said.
9(MDAxODcyMTgzMDEyMTgxMTY5NjYxN2I4OQ004))
Transcript
LEILA FADEL, HOST:
The White House has been saying for months that it wants to manage both the risks and rewards of artificial intelligence.
MICHEL MARTIN, HOST:
Now major tech companies working with the White House have made voluntary commitments on how they will develop, test and share AI systems. The president and company leaders will be speaking about what they've agreed to this afternoon.
FADEL: NPR's Deepa Shivaram covers the White House and joins us now. Hi, Deepa.
DEEPA SHIVARAM, BYLINE: Hey. Good morning.
FADEL: Good morning. So which tech companies are involved here? And what's in these agreements?
SHIVARAM: Yeah, there are seven tech companies that have agreed to these commitments, and those include Amazon, Anthropic, Google, Inflection, Microsoft Meta and OpenAI, which is the parent company of ChatGPT. The leaders of these seven companies will be at the White House today to talk about these agreements, which are essentially parameters for how they'll develop AI technology and roll it out for public use.
So for example, the companies committed to making sure users know when content is AI-generated through something like a watermark. They also say they'll make a point to avoid bias in their technology and protect privacy. But overall, there aren't a lot of details in what the White House has released today. So at this point, it's hard to say how effective these commitments will be or if more companies will choose to join in on these agreements.
FADEL: Well, speaking about effectiveness, I mean, these commitments are voluntary, right? So how will the companies be held accountable?
SHIVARAM: Yeah, that's definitely a concern for a lot of people. There are a number of polls that show public trust in big tech companies to do the right thing is pretty low. I asked Jeff Zients, the White House chief of staff, about this. And he says the federal government will do whatever they can to hold these companies to their commitments, but it's just a first step.
JEFF ZIENTS: Commitments the companies are making are a good start, but it's just a start. And the key here is implementation and execution. In order for these companies to perform and earn the public's trust.
SHIVARAM: And he pointed specifically to one of the commitments, which is to have external checks on emerging AI technology from independent third parties. For example, some of the companies will have their AI systems tested at a hacking conference next month at the encouragement of the White House. But beyond that, we don't really have much detail on who serves as these third-party checks on the technology and how those people are selected.
FADEL: And there really is so much concern about how this technology is going to be regulated because it could have so much impact on how people work, what's true, what's not true. So we're waiting on details on these commitments. But what other actions are we anticipating coming from the White House in the meantime?
SHIVARAM: Right. In the next few weeks, the White House is planning to release executive actions on AI. The Office of Management and Budget will also be releasing guidance to federal agencies on how they can and cannot use AI in government work. And the White House says they're also closely working with Congress as they develop legislation on regulating AI. And on top of what's happening in the U.S., there's also a lot of global conversations happening, too.
The White House says they've consulted on the agreements they announced today with dozens of other countries. And Biden continues to speak about AI with foreign leaders. He mentioned AI at the Nordic Leaders' Summit last week in Helsinki. And he and U.K. Prime Minister Rishi Sunak have been talking about AI with each other consistently. It's a lot of talk, so what concrete actions the White House takes next on the international front will also be something to watch.
FADEL: That's NPR's Deepa Shivaram. Thanks so much, Deepa.
SHIVARAM: Thank you. Transcript provided by NPR, Copyright NPR.
300x250 Ad
300x250 Ad