Meta Says Limits on Sharing AI Technology May Dim US Influence
(Bloomberg) -- Meta Platforms Inc. policy chief Nick Clegg warned US government officials that limiting companies from sharing certain artificial intelligence technologies publicly could harm American influence over the relatively new and evolving global AI industry.
Most Read from Bloomberg
Trump Sues Truth Social Company Co-Founders to Zero Them Out
A Million Simulations, One Verdict for US Economy: Debt Danger Ahead
Trump Got His $175 Million Bond From a Billionaire Fan’s Company
In public comments submitted Wednesday to the Commerce Department’s National Telecommunications and Information Administration, Meta wrote that a “restrictive approach” to sharing AI models may mean that technologies developed in other countries “become the global norm.”
“It’s not as if that vacuum is not going to be filled by others,” Clegg, Meta’s president of global affairs, said in an interview. “Open-sourcing is actually the way you export US values and technology.” The public input from Meta, the parent of social media apps Facebook, Instagram and WhatsApp, was shared with the White House after an open request for comment last month by the Biden administration, which is considering how to regulate AI technology.
Major players in the AI industry disagree about whether the programming and data behind artificial intelligence models should be public and open, or held more closely as proprietary information. Startup OpenAI, one of the industry’s early leaders, has been in favor of keeping its technology private, while Meta and some other major tech giants, like GitHub owner Microsoft Corp., have been in favor of open-sourcing at least some AI code.
Meta’s argument for keeping the technology open may resonate with the current presidential administration and Congress. For years, efforts to pass legislation that restricts technology companies’ business practices have all died in Congress, including bills meant to protect children on social media, to limit tech giants from unfairly boosting their own products, and to safeguard users’ data online.
But other bills aimed at protecting American business interests have had more success, including the Chips and Science Act, passed in 2022 to support US chipmakers while addressing national security concerns around semiconductor manufacturing. Another bill targeting Chinese tech giant ByteDance Ltd. and its popular social network, TikTok, is awaiting a vote in the Senate after passing in the House earlier this month.
Meta, based in Menlo Park, California, has been one of the technology industry’s most vocal proponents of federal AI regulation, including rules to protect against deepfakes. Clegg said he’s optimistic that the government will act to develop rules around AI, despite lawmakers’ failure to agree on many other issues.
The technology industry has been largely unregulated since Meta’s founding in 2004 — a reality that has led to questions about whether the biggest companies have been allowed to amass too much power. “I think there’s sort of a feeling of like, ‘Let’s not let this happen again,’” Clegg said.
Most Read from Bloomberg Businessweek
©2024 Bloomberg L.P.
