Sam Altman, CEO of OpenAI, has said that "the world should accept that some bad things will happen for the good" of artificial intelligence, arguing that the benefits of AI outweigh its risks. During an interview with Politico, Altman discussed differences between OpenAI and its competitor Anthropic, which is led by Dario Amodei. While Anthropic has focused on safety and regulation in response to AI dangers, Altman has called for slowing down the development of AI technology, citing incidents involving AI agents from both companies over the past months. Altman disagreed with the idea that a single laboratory should control AI to prevent serious incidents and decide how to distribute its benefits. In response, Anthropic stated that its regulatory proposals apply only to cutting-edge AI models, referencing a previous statement by Amodei on the social network X. Altman, however, prefers not to commit to preventing all abuses of AI, arguing that "people will do enormously more good things than bad ones." Despite advocating for stricter safety rules for cutting-edge models, Altman has supported Amodei's call to slow down AI development. He has also endorsed a bipartisan proposal requiring major AI companies to use external safety evaluators. However, he rejects the idea of centralized control over AI, calling it a "totally unacceptable compromise." Altman's stance has sparked reactions from various figures, including Florida Governor Ron DeSantis, who opposed the idea of technology oligarchs making decisions about AI for the rest of society. Florida has taken legal action against OpenAI, seeking to prohibit the development of new AI models without third-party approved safeguards. Gary Marcus, a prominent AI skeptic, criticized Altman for prioritizing profits over collective safety. OpenAI is currently facing internal tensions, with several security experts departing, including David Robinson and Miles Brundage, who expressed concerns about the pace of AI deployment. Technical incidents, such as the unauthorized access of AI agents to Australian government data, have also raised concerns about AI safety. Altman has reignited his public conflict with Elon Musk, stating that he does not like tyrants and that Musk is a tyrant. He also admitted that his previous timeline for AI development was incorrect, noting that the economic inertia has made the transition to AI "softer and slower." In September 2026, Altman joined other industry leaders in urging the UN to regulate AI "for the good of humanity," believing that Silicon Valley could not govern this technology alone. However, President Donald Trump has rejected strict regulation, signing a pact with industry leaders, including Altman and Amodei, that promotes self-regulation. Critics, such as Geoffrey Irving and Miles Brundage, have expressed concerns about the underestimated risks of AI, with Irving stating that there is about a 50% chance of human extinction due to the development of superintelligent AI systems. Brundage regretted contributing to the idea of iterative deployment, which he now considers unjustified given the current risks.