OpenAI open to slowing AI race as fears over losing control mount: report

OpenAI CEO Sam Altman told employees at an internal company meeting this week that OpenAI is open to slowing development of its AI systems and coordinating the pace with rival AI labs, Bloomberg News reported. The disclosure follows OpenAI's two-week pause of much of its model development in August to strengthen safeguards after AI agents broke containment and hacked the open-source platform Hugging Face, and comes as Anthropic scientist Evan Hubinger put the chance of AI killing all humans within the next decade at greater than 10 percent. OpenAI said Wednesday it is lobbying for mandatory federal AI safety requirements, while President Donald Trump argued the more immediate priority is ensuring the United States beats China in the global AI race.

OpenAI open to slowing AI race as fears over losing control mount: report OpenAI Chief Executive Sam Altman told employees at a company meeting this week that the firm was open to slowing development of its AI systems - Bookmark OpenAI CEO Sam Altman /news/world/americas/us-politics/openai-utility-security-sam-altman-takeover-b3048287.html has reportedly told employees the company is open to slowing the development of its artificial intelligence systems amid escalating concerns over the risks posed by increasingly powerful models. Altman raised the possibility during an internal company meeting this week, Bloomberg News reported, citing people familiar with the discussion. The potential move would involve coordinating the pace of development with rival AI labs as researchers and lawmakers increasingly grapple with how to keep advanced systems under control. OpenAI /news/openai-josh-hawley-washington-chris-van-hollen-bernie-sanders-b3048313.html did not comment on Altman’s reported remarks. The company has previously acknowledged that the accelerating capabilities of frontier AI could eventually require developers to “pace the rate of AI advancement.” Those concerns intensified in August, when OpenAI temporarily paused much of its model development for two weeks to strengthen safeguards after AI agents broke containment and hacked the open-source platform Hugging Face. Warnings have also emerged from researchers working inside some of the world’s leading AI companies. Jacob Coxon, a former researcher at OpenAI and Anthropic who recently resigned, accused developers this week of racing toward increasingly powerful technology without taking sufficient precautions. “The people building AI earnestly believe that it could kill us all by the end of the decade,” Coxon said. Anthropic scientist Evan Hubinger backed that warning, writing on X: “We really do earnestly believe AI could kill all humans.” He said he personally puts the chance of that happening within the next decade at greater than 10 percent. Anthropic said Thursday it was interested in working with other companies on the timing of new technology releases while continuing rigorous testing of models and potentially dangerous capabilities, including those involving cybersecurity and biology. OpenAI, meanwhile, said Wednesday it is lobbying for mandatory federal AI safety requirements amid concerns that sufficiently advanced systems could eventually accelerate their own development. The warnings are beginning to attract attention in Washington. Democratic Sen. Mark Kelly of Arizona said the federal government needed to “wake up and take this seriously,” while Republican Sen. Ted Cruz of Texas has pointed to potential legislation addressing “catastrophic risks” from AI. Republican Rep. Anna Paulina Luna of Florida has urged House Speaker Mike Johnson to convene a special session of Congress devoted to the technology. President Donald Trump, however, has pushed back against fears that AI could pose an existential threat, arguing that the more immediate priority is ensuring the United States beats China in the global race for AI supremacy. “I have concerns that if we don't win AI, we're going to be put in a very bad position,” Trump told reporters at Dallas Fort Worth International Airport. “We are leading China right now by a pretty good period.” Other researchers have left leading AI companies over safety concerns. Josh Engels, a former Google DeepMind safety researcher who has since joined the nonprofit METR, told NBC that “there are no adults in the room,” while former Anthropic safety team lead Joe Benton warned that AI could advance too quickly for governments and developers to respond effectively. Lawmakers are now considering several proposals to address those concerns. Senate Majority Leader John Thune and Democratic Sen. Amy Klobuchar are working on legislation targeting potentially catastrophic AI risks, while a bipartisan group of six House lawmakers has proposed requiring independent security audits accredited by the Commerce Department. California Gov. Gavin Newsom has also signed the first state law establishing third-party evaluation standards for AI products. Alicia Solow-Niederman, a technology law professor at George Washington University, said increasingly capable systems require room for outside scrutiny. “It is important to have space for oversight of these systems, including appropriately calibrated pre-deployment review and independent third-party audits,” she said. Join our commenting forum Join thought-provoking conversations, follow other Independent readers and see their replies Comments comments-area