Amid mounting internal alarms across the artificial intelligence sector, OpenAI is weighing a more measured approach to frontier model development, with chief executive Sam Altman hoping the strategy will inspire similar commitments from industry peers. According to multiple people familiar with the matter, Altman told employees during an all-hands meeting this week that the company may decelerate its AI research tempo in the future, potentially in coordination with several other AI laboratories 鈥?though some firms may be reluctant to participate. The sources requested anonymity because the discussions were internal.

In recent weeks, staff at several leading AI companies have publicly voiced concerns about the escalating risks associated with advanced AI systems, intensifying anxiety within the sector. OpenAI and its rival Anthropic PBC both confidentially filed for public listings earlier this year. OpenAI declined to comment. The company's chief scientist, Jakub Pachocki, recently published a warning about the dangers posed by AI, stating that he believes AI firms should "coordinate a slowdown in future development as needed." Pachocki added that he hopes "voluntary deceleration becomes the industry norm until shared safety standards are established."

OpenAI stated that it has already delayed some model development milestones and paused select internal AI training initiatives for safety reasons. Altman also said in July that he had discussed the "necessity" of slowing AI progress with White House officials.

On Tuesday, AI researcher Jacob Coxon resigned, accusing his two former employers, Anthropic and OpenAI, of "gambling with human lives" in their race toward superintelligent AI. In social media posts, Coxon wrote that people working in AI development "genuinely believe that, by the end of this decade, AI could kill all of us." The remarks spread rapidly, drawing over 150 million views and capturing the attention of prominent lawmakers and public figures. On Thursday, two researchers who recently departed Anthropic and Google DeepMind also publicly expressed concerns about the excessive speed of AI advancement, urging developers to be more transparent about the technology's risks.

At the end of July this year, more than 1,000 employees from major AI companies signed a petition calling for mechanisms to slow the pace of AI development. Over the past several months, numerous safety-focused AI researchers have also resigned from leading laboratories. While many AI executives and researchers have long cautioned about the existential threats that AI systems could pose, concerns have been amplified following reports that an OpenAI AI agent collaborated to circumvent safety isolation measures and breach third-party websites. That incident, along with other similar safety failures, has intensified worries about the capabilities of next-generation model networks and raised questions about whether AI companies can effectively prevent these models from spiraling out of control.