OpenAI is considering slowing the development of increasingly powerful artificial intelligence models as concerns grow over the risks posed by advanced AI, according to a report by Bloomberg.
OpenAI CEO Sam Altman discussed the possibility during a company-wide meeting, saying the company could put further AI advancement on hold while it works to better understand and contain the risks associated with more capable models.
The proposal comes as AI systems become increasingly capable of carrying out complex tasks, including activities that could potentially be misused for cyberattacks and other harmful purposes.
OpenAI wants rivals to slow down too
Altman reportedly expressed hope that other leading AI companies would also slow their development rather than use a pause by OpenAI as an opportunity to move ahead.
The idea reflects growing concern within the AI industry that competition to develop increasingly capable models could outpace the development of adequate safety measures.
OpenAI chief scientist Jakub Pachocki has also called for greater coordination among AI developers.
Pachocki said AI companies should consider “coordinating to slow down future development as needed” and expressed hope that voluntary slowdowns would become common until shared safety standards are established.
Would other AI companies agree?
A major challenge would be persuading competing AI developers to adopt a similar approach.
Anthropic has also faced internal concerns over the potential risks of increasingly capable AI systems. One of its researchers recently raised warnings about the possibility of AI posing an existential threat to humanity.
The issue is further complicated by competition beyond the major US-based AI companies.
A coordinated slowdown among leading American firms could potentially give overseas developers an opportunity to advance their own models more quickly, raising concerns about a new phase of the global AI race.
The emergence of China’s DeepSeek earlier highlighted how quickly a relatively unexpected competitor can disrupt assumptions about the balance of power in artificial intelligence.
AI safety debate intensifies
OpenAI’s reported willingness to consider a slowdown comes as governments, researchers and AI companies face increasing pressure to establish stronger safeguards for frontier models.
The debate is no longer focused only on the future capabilities of AI but also on how quickly those capabilities are advancing and whether safety systems can keep pace.
For OpenAI, any voluntary pause would therefore involve balancing the risks of moving too quickly against the possibility that competitors could continue developing more powerful systems.
