Sam Altman's Warning: Accept Risks for AI Benefits

OpenAI chief executive Sam Altman says the world should be willing to accept "some bad things" from artificial intelligence, because he believes the good the technology does will far outweigh the harm. His comments, made in an interview with Politico and reported by Forbes on October 4, 2026, quickly spread online because they put a blunt trade-off at the center of the AI safety debate.

What Altman said In the interview, Altman argued that people will use AI to do "orders of magnitude more" good things than bad things. In his view, that balance justifies accepting some harm along the way rather than holding the technology back. He also said he does not believe warnings about "really catastrophic risks" from AI, including fears that humans could suffer "a serious loss of control" over the systems they build.

A split with Anthropic Altman used the interview to draw a sharp line between OpenAI and rival AI company Anthropic, saying there is "broad daylight" between the two. He disagreed with the idea that a "single lab in San Francisco" should be in charge of controlling AI in order to prevent harm. According to the Forbes report, OpenAI favors a lighter-touch approach to regulation, while Anthropic has taken a more cautious stance.

Who is involved OpenAI is the company behind ChatGPT, the chatbot that brought AI assistants into the mainstream after its launch in late 2022. Altman has led the company through its rapid growth and is one of the most closely watched voices in the industry, so his views on risk and regulation carry weight with lawmakers, investors and other tech companies. Anthropic, the company he contrasted OpenAI with, makes the Claude family of AI models and was founded by former OpenAI employees.

Why it matters Governments are still deciding how tightly to regulate advanced AI, and companies are lobbying for very different approaches. A leading CEO saying openly that society should accept some harm in exchange for the benefits is likely to sharpen that debate. Supporters will see it as an honest description of how every powerful technology arrives, with benefits and costs together. Critics will ask who gets to decide which harms are acceptable, and who pays for them. How that question is answered will shape the rules that govern AI tools millions of people already use every day.