Search

Biden plans to strengthen government oversight of AI with new "pressure testing" 2024-04-17 05:11:58

On Monday, President Biden announced significant executive actions aimed at trying to establish oversight over the rapidly advancing artificial intelligence sector by setting new standards for product safety testing and creating a system of federal trials for critical systems.

The order, which had been in the works for several months, reflects the White House's concern that leaving technology unchecked could pose significant risks to national security, the economy, public health, and privacy. This announcement comes just days before a major global AI summit in London, which Vice President Harris will attend.

"In the next 10, maybe in the next five years, we'll see more technological change than we've seen in the last 50 years," Biden said during a speech at the White House.

"The most important technology of our time, artificial intelligence, is accelerating these changes," Biden added.

President Biden at a meeting with AI company leaders at the White House on July 21. The White House initially worked on voluntary commitments, but the new actions go beyond that. Anna Moneymaker / Getty Images The order requires companies to share the results of testing for key AI systems. Tech companies currently conduct their own "red teaming" of products, subjecting them to testing to identify potential issues like disinformation or bias. The White House has already worked with major developers on a set of voluntary commitments to vet their systems with third parties before release.

However, Biden's order calls on the government to establish new standards, tools, and tests for "red teaming" and requires companies to notify the government and share the results of their "red teaming" for products that may pose serious risks before deploying the systems. The authority to require companies to do this stems from the Defense Production Act, a Korean War-era law that expands presidential powers, particularly concerning national security issues.

"We need to manage this technology. In my view, there's no other way around it. It needs to be managed," Biden said.

Jeff Zients, White House Chief of Staff, speaks with Senate Majority Leader Chuck Schumer at a White House event focused on artificial intelligence on October 30. Chip Somodevilla / Getty Images In an interview with NPR, White House Chief of Staff Jeff Zients said the order introduces "pressure testing" into the AI development process.

"At the end of the day, companies can't self-assess here," Zients said. "So, we've set new standards for engaging with the private sector on AI, and we're going to ensure that private companies adhere to them."

While the Defense Production Act allows the White House to have some authority over new product development by companies, there is still an overall lack of regulatory enforcement, which, according to White House officials, will require congressional legislation.

Congress wants to regulate AI, but it has a long way to go POLITICS Congress wants to regulate AI, but it has a long way to go Biden announced that a bipartisan group of senators, led by Senate Majority Leader Chuck Schumer, will meet with him at the White House on Tuesday to discuss further actions that Congress can take to develop new AI laws.

The White House sees opportunities in AI but also warns about risks Biden said he sees the potential of AI in developing new drugs and cancer treatments. But the White House also expressed concerns about shortcomings, such as "elderly people being deceived by voice cloning technology," said Zients.

"AI can use data - your personal data - to make social media even more addictive for you or your children. That's not good. AI systems can use your data to discriminate against a person of color trying to buy a home. That's unacceptable," Zients said.

The new order requires agencies to explore established standards for how departments can use AI for positive purposes, such as developing new medicines and aiding teachers in classrooms. However, it also calls on agencies to establish standards to ensure that AI does not exacerbate discriminatory practices in housing or lending decisions.

But some experts, like Ifeoma Ajunwa of Emory University, say that the order doesn't pay enough attention to everyday AI risks.

"The real current danger is not that AI is becoming too smart. Instead, people are using AI in ways that are inconsistent with our democratic beliefs in equal opportunity and equal protection," Ajunwa told NPR.

"It has no concrete power over mandatory AI hiring practices to ensure that they don't have a discriminatory effect," she added, noting that using AI in hiring is already a standard practice.

Science

Unraveling Diplomatic Threads: Mike Johnson's Bold Move to Diverge Israel and Ukraine Aid
Roku Security Breach: 576,000 Streaming Accounts Compromised, Company Reports
Transcript Excerpt: IMF Director Kristalina Georgieva on "Face the Nation" - April 14, 2024
Historic Triumph: Gymnast Morgan Price Makes History as First HBCU Athlete to Claim National Collegiate Title
Tax Trap Alert: Beware of Viral Tax Tips on TikTok, Caution Experts

Contact

Contact us: [email protected]

Welcome to "The Stream News," your premier source for up-to-the-minute, reliable, and insightful news coverage from around the world. With a commitment to delivering accurate and diverse news stories, we aim to keep you informed, engaged, and enlightened on the most pressing issues and captivating events shaping our global community.