Monday, January 29, 2024

AI companies will need to start reporting their safety tests to the US government

Mon, January 29, 2024 

WASHINGTON (AP) — The Biden administration will start implementing a new requirement for the developers of major artificial intelligence systems to disclose their safety test results to the government.

The White House AI Council is scheduled to meet Monday to review progress made on the executive order that President Joe Biden signed three months ago to manage the fast-evolving technology.

Chief among the 90-day goals from the order was a mandate under the Defense Production Act that AI companies share vital information with the Commerce Department, including safety tests.

Ben Buchanan, the White House special adviser on AI, said in an interview that the government wants "to know AI systems are safe before they’re released to the public — the president has been very clear that companies need to meet that bar.”

The software companies are committed to a set of categories for the safety tests, but companies do not yet have to comply with a common standard on the tests. The government's National Institute of Standards and Technology will develop a uniform framework for assessing safety, as part of the order Biden signed in October.

AI has emerged as a leading economic and national security consideration for the federal government, given the investments and uncertainties caused by the launch of new AI tools such as ChatGPT that can generate text, images and sounds. The Biden administration also is looking at congressional legislation and working with other countries and the European Union on rules for managing the technology.

The Commerce Department has developed a draft rule on U.S. cloud companies that provide servers to foreign AI developers.

Nine federal agencies, including the departments of Defense, Transportation, Treasury and Health and Human Services, have completed risk assessments regarding AI's use in critical national infrastructure such as the electric grid.

The government also has scaled up the hiring of AI experts and data scientists at federal agencies.

“We know that AI has transformative effects and potential,” Buchanan said. “We’re not trying to upend the apple cart there, but we are trying to make sure the regulators are prepared to manage this technology.”

Josh Boak, The Associated Press


White House AI council meets Monday as legislative action stalls


Mon, January 29, 2024 



WASHINGTON (Reuters) - The White House artificial intelligence council is meeting Monday, three months after President Joe Biden signed an executive order that aims to reduce the risks AI poses.

White House Deputy Chief of Staff Bruce Reed, who will convene the council meeting Monday, said in a statement the federal government had made significant progress in the prior 90 days on AI, saying Biden's "directive to us is move fast and fix things."

The White House said nine government agencies - including Defense, Transportation, Treasury, and Health and Human Services - submitted risk assessments to the Department of Homeland Security required under Biden's order.

At the same time, efforts in Congress to pass legislation in Congress addressing AI have stalled despite numerous high-level forums and legislative proposals.

On Friday, the Biden administration said it was proposing requiring U.S. cloud companies to determine whether foreign entities are accessing U.S. data centers to train AI models through "know your customer" rules.

"We can't have non-state actors or China or folks who we don’t want accessing our cloud to train their models," U.S. Commerce Secretary Gina Raimondo told Reuters Friday. "We use export controls on chips. Those chips are in American cloud data centers so we also have to think about closing down that avenue for potential malicious activity."

Last month, Raimondo said Commerce would not allow Nvidia "to ship is the most sophisticated, highest-processing-power AI chips, which would enable China to train their frontier models."

Biden's executive order invokes the Defense Production Act to require developers of AI systems that pose risks to U.S. national security, the economy, public health or safety to share the results of safety tests with the U.S. government before they are publicly released.

The Commerce Department plans to soon send those survey requests to companies. Raimondo told Reuters companies will have 30 days to respond.

"Any company that doesn't want to comply is a red flag for me," she said.

Top cloud providers include Amazon.com's AWS, Alphabet's Google Cloud and Microsoft's Azure unit.

(Reporting by David Shepardson; Editing by Stephen Coates)


US Wants Cloud Firms to Reveal Foreign Clients in China AI Race

Courtney Rozen and Mackenzie Hawkins
Sun, January 28, 2024 


(Bloomberg) -- The US wants cloud services providers such as Amazon.com Inc. and Microsoft Corp. to actively investigate and call out foreign clients developing artificial intelligence applications on their platforms, escalating a tech conflict between Washington and Beijing.

The Biden administration proposal, scheduled for release Monday, requires such firms to reveal foreign customers’ names and IP addresses. Amazon and its peers, which include Alphabet Inc.’s Google, would have to devise a budget for collecting those details and report any suspicious activity, according to draft rule published Sunday.

If implemented, Washington could use those requirements to choke off a major avenue through which Chinese firms access the data centers and servers crucial to training and hosting AI. They also place the onus of collecting, storing and analyzing customer data on the cloud services, a burden not unlike strict “know-your-customer” rules that govern the financial industry. US cloud providers have worried that restrictions on their activities with overseas users without comparable measures by allied countries risks disadvantaging American firms.

Representatives for Microsoft, Amazon and Google didn’t immediately respond to requests for comment outside normal US hours. A Commerce Department spokesperson referred Bloomberg to Commerce Secretary Gina Raimondo’s comments last week.

Raimondo said Friday her team was working to eradicate national security threats posed by AI development, an effort likely to focus on firms from China. Washington, which has already worked to constrain Beijing’s access to the most advanced semiconductors, wants to limit Chinese firms’ ability to develop AI with potential military capabilities.

“These models getting in the hands of non-state actors or people that aren’t our allies is very dangerous,” Raimondo said in Washington.

President Joe Biden in October directed the Commerce Department to require such disclosures in an effort to detect foreign actors that might use AI to launch what the proposal dubs “malicious cyber-enabled activities.”

The US is asking for comments on the proposed rule through April 29 before finalizing the regulation.

Commerce said it may provide an exception to the identification rules for the foreign subsidiaries of US cloud providers. It also referred to commenters so far who’ve pushed for the broadest possible definition of a US cloud service, adding it will clarify whether foreign subsidiaries fall under the rules.

China’s development of AI and other next-generation technologies is a top concern for the administration, which sees Beijing as its primary global strategic competitor.

Washington has tried to rein in China’s advances by restricting chip exports to the country and sanctioning individual Chinese firms, but the country’s tech leaders have managed to make significant breakthroughs despite US curbs.

The US in October tightened its controls to capture more chips, equipment and geographies. One key update targeted Chinese-headquartered companies operating in more than 40 countries, an attempt to prevent those firms from using other nations as intermediaries to secure semiconductors they can’t access at home.

 Bloomberg Businessweek


No comments:

Post a Comment