Sam Altman believes the world should accept the “bad things” that happen with AI in exchange for the benefits of the technology.
OpenAI’s chief executive cited hacks, scams and “other bad things that are going to happen” in an interview that sparked immediate backlash from critics of big AI companies. His comments come after one of his company’s security experts resigned, saying over the weekend that its “culture is broken.”
In an interview published Monday, Altman was asked about the difference between OpenAI’s approach to security and that of Anthropic, which called on the industry to slow down after the departure of one of its researchers, warning that AI experts believed “this could kill us all by the end of the decade.”
“I think there’s a lot of daylight,” Altman said. “One of the differences between us and some of the strictest AI safety experts is that we think the world should accept that some bad things happen for the benefits of this technology and for the people with the agency (to use AI widely).
“I think the lighter regulatory stance that we’re advocating comes with an acceptance that bad things are going to happen as society understands its resilience.
“I wouldn’t agree to say that we will make sure that there is no major hack, that there is no misuse of this technology, that there are no scams or all the other bad things that will happen because I think people will do enormously – orders of magnitude – more good things than bad things.”
The interview for Politico’s Decoded podcast drew a strong reaction from Ron DeSantis, the governor of Florida. He said he had “no chance” with the idea “that a handful of tech oligarchs can make this (security) decision for the rest of us.”
His state asked a judge last week to bar OpenAI from developing new AI models without third-party approved guardrails.
Gary Marcus, a prominent AI skeptic and professor emeritus at New York University, said Altman was “saying the silent part out loud: Suck it up, so you can make us rich and powerful.”
Altman also reignited a feud with Elon Musk, telling Vanity Fair: “I don’t like bullies, and I think Elon is a bully.”
He was asked if the AI industry could “grow up” and work together.
“I think every once in a while you have to punch a bully in the nose,” Altman said. “And I also think that Elon, like a lot of bullies, only respects people who do it. So if you want me to be able to work with him, maybe you should be happy for me to fight back from time to time, but I’ll work with Elon, of course.”
Altman co-founded OpenAI with Musk in 2015, but earlier this year, Musk alleged in court that Altman, OpenAI and the company’s chairman, Greg Brockman, had unjustly enriched themselves and breached a founding contract. Musk’s request was rejected by a judge and jury.
OpenAI was hit by multiple security crises over the summer, including when a swarm of its AI agents escaped from their training “sandbox,” cheated, deceived and conspired to hack the Hugging Face website. Other agents accessed Australian government data. In the Vanity Fair interview, Altman warned of “a coming tidal wave of cybersecurity problems” from competing open source models. He suggested that OpenAI could help its customers defend against this risk.
after newsletter promotion
The company has backed away from releasing a cutting-edge model, but Altman’s comments reflect a more optimistic mood as an IPO approaches.
Last month’s wave of public and political concern over the potentially catastrophic consequences of rapidly advancing AI agents that act beyond human intent has raised hopes for U.S.-wide and even international security agreements.
However, Donald Trump and AI bosses, including Altman and Anthropic’s Dario Amodei, signed a much more laissez-faire pact at the White House last week. The US president said he wants companies to “police themselves” to limit risks from cyberattacks and AI-based biological weapons.
Geoffrey Irving, who has worked at OpenAI and Google DeepMind and as chief scientist at the UK’s AI Safety Institute, said warnings about AI risks underestimated the dangers.
“There is about a 50% chance that we will all die from the development of AI systems that are smarter than humans,” he said, adding that actions taken over the next two to 10 years would be decisive.
OpenAI’s David Robinson announced his departure over the weekend, saying “the companies building this technology aren’t careful enough” when sprinting from one product launch to the next.
His view was supported by Miles Brundage, an AI policy researcher who worked at OpenAI for six years. He said he regretted helping “spread the idea of iterative deployment,” which might have made sense when AI was less powerful, “but makes no sense after so many deaths have been linked to AI and as we move toward extinction risks.”
Gn bussni
