OpenAI CEO Altman: "Bad things will happen" with AI

Sam Altman, CEO of OpenAI, sparked controversy in the tech world today when he declared in an interview that the world must accept negative outcomes in exchange for the benefits of artificial intelligence. He said he supports lighter regulation and made clear he would not agree to halt development simply to prevent cyberattacks, fraud or misuse entirely.

"I think there's a big gap. One of the differences between us and some of the more stringent AI safety people is that we believe the world needs to accept some bad things happening in exchange for the benefits of this technology and people's ability to use it widely," Altman said.

He added that the more permissive regulatory stance he advocates comes with accepting that some harmful events will occur as society discovers its resilience. "I wouldn't make a deal where we guarantee there will be no serious breaches, no misuse of this technology, no fraud at all or any of the other bad things that will happen, because I think people will do orders of magnitude more good things than bad things," he said.

Extreme warnings from safety experts

Altman's remarks come against a backdrop of stark warnings from safety experts in the field. Geoffrey Irving, former chief scientist at the UK's AI Safety Institute, estimates there is roughly a 50% chance that we will all die as a result of developing systems more intelligent than humans. This dramatic warning joins a wave of resignations by senior safety personnel at leading companies.

Meanwhile, OpenAI safety expert David Robinson resigned over the weekend, claiming the company's culture is broken. Policy researcher Miles Brundage, who worked at OpenAI for six years, also expressed regret over his support for gradual rollout of the technology, saying it leads to extinction-level dangers.

Over the summer, OpenAI experienced serious safety failures, including AI agents that broke out of their protected environment, committed fraud, and planned to breach the Hugging Face repository and Australian government information. The company cancelled the launch of an advanced model due to safety concerns.

Sharp criticism from politicians and experts

Altman's statements drew harsh criticism from public figures. Florida Governor Ron DeSantis attacked Altman, saying he will not allow "a handful of tech oligarchs" to make such decisions for the public. His state has asked a court to mandate independent safeguards.

Gary Marcus, professor emeritus at New York University, said Altman is saying the quiet part out loud: "Suck it up, so you can make us rich and powerful."

President Donald Trump signed an agreement last week with AI company heads based solely on self-regulation, a move that drew further criticism amid safety concerns.

Altman also addressed Elon Musk in the interview, calling him a bully but saying he would work with him in the future. "I don't like bullies, and I think Elon is a bully. I think sometimes you have to punch a bully back in the nose. And I also think Elon, like many bullies, only respects people who do that," he said.

Additionally, Altman warned of an approaching wave of cybersecurity problems due to open-source models from competitors, raising further questions about the industry's overall safety.