OpenAI’s Chris Lehane thinks the discussion around AI has gotten out of hand. 

“Some of the conversation out there is not necessarily responsible,” he told The Standard on Tuesday. “And when you put some of those thoughts and ideas out there, they do have consequences.”

“This is not fun and games,” he continued. “This is really serious shit.” 

Even before a 20-year-old from Texas traveled to San Francisco to lob a Molotov cocktail at the gate of OpenAI CEO Sam Altman’s San Francisco home, Lehane, who joined OpenAI as global policy chief in August 2024, was trying to change the narrative on AI. 

Lehane is a D.C. veteran and former crypto evangelist who has been helping OpenAI tell the story of how its technology can improve the lives of everyday Americans.

Friday’s attack and the public response to it shows that Lehane faces an uphill battle. 

“Our job at OpenAI and in the AI space — and we need to do a much better job — is to explain to people why … this is going to be really good for them, for their families and for society writ large,” he said.

• The AI IPOs are coming. What will be the cost for San Francisco? 

• What the pope’s AI manifesto means for Silicon Valley

• Case dismissed: Jury rules Musk missed his chance

Lehane is up against legitimate economic and ethical concerns, as well as a growing and fractured movement of anti-AI thinkers and activists. Daniel Moreno-Gama, the man charged with attempted murder for throwing a homemade bomb at Altman’s house, appears to be in the latter group. 

Posts in a Substack account bearing Moreno-Gama’s name suggest he has existential fears about the rise of AI. He appeared to be a member of PauseAI, an international group that opposes AI development via nonviolent protest. In a statement asking for privacy, Moreno-Gama’s family said he had been experiencing a “mental health crisis” prior to attacking Altman’s home. 

PauseAI acknowledged that Moreno-Gama had been a member of its Discord community for two years, but said in a statement (opens in new tab) Sunday that none of the 34 messages he posted during that time contained violent language. The group said some but not all of those messages were deleted in the process of banning Moreno-Gama’s account. 

Lehane said the group’s move to erase messages was “telling.” 

Have thoughts on this story?

That people are worried about AI is understandable, Lehane said — they believe it might take their jobs, harm their kids, and raise their electricity bills. He compared the tension to conflicts that followed earlier technological leaps forward, like the invention of the printing press. And it doesn’t help, he said, that the AI industry has made a habit of foreboding pronouncements. 

“You’ve had a series of things that have been put out there — but haven’t come to fruition — about extreme things that are going to happen,” Lehane said.

Just last week, Anthropic announced that it was holding back the release of its latest model, Mythos, because of concerns about the havoc it could wreck on the cybersecurity industry and, consequently, geopolitics. Instead of releasing the model to the public, the company said it would make it available  to a coalition of partner companies to review their internal vulnerabilities, in a process it’s calling “Project Glasswing.” 

Lehane wants people to believe that AI can enhance their lives — whether by improving medical treatment or making their jobs easier. But he said the conversation has become too binary. 

“You have one group that effectively says, ‘This is going to be the greatest thing ever, everyone’s going to be living in beachside homes, painting in watercolors as they while away their days.’ And then you have another extreme, which I would call the Doomers, who have a very, very negative and dark view of humanity.” 

Neither group, Lehane said, offers much in the way of a solution to the very real problems AI presents. An OpenAI white paper published last week, which Lehane said the company has been working on since last year, delves into ways he thinks AI can “create incredible economic opportunities” for people beyond the tech industry. These include an enhanced social safety net and worker-led organizations that provide entrepreneurs with AI tools and skills. 

Some of these ideas echoed previous proposals. When OpenAI announced its ambitious AI infrastructure plan last year, it said it would create economic development zones (opens in new tab) and give local governments free access to AI resources. So far, that hasn’t happened.

Lehane says research has shown that the more people use AI, the more they believe it can have a positive impact on their lives, with power users being the most bullish of all. Outside the U.S. and Europe, people tend to be much more optimistic about AI, he said.

“What’s the old FDR line?” he said. “There’s nothing to fear but fear itself.”