Neil deGrasse Tyson says AI needs guardrails, not a shutdown: ‘Don’t put it all in one bin’
TheGrio...
attends the US premiere of DISCLOSURE DAY presented by Universal Pictures and Amblin Entertainment on June 08, 2026 in New York City. (Photo by Bryan Bedder/Getty Images for Universal Pictures and Amblin Entertainment )
Tyson says potentially dangerous AI applications need guardrails while pointing to the technology’s benefits in medicine and science.
Neil deGrasse Tyson is acknowledging the potential dangers of artificial intelligence while pushing back on the idea that all AI should be treated as an existential threat.
The famed astrophysicist spoke with TMZ about the future of AI days after former Anthropic researcher Jacob Coxon publicly warned that the technology could eventually threaten humanity. TMZ spoke with Coxon recently about his decision to leave Anthropic, where he had worked on AI research.
Coxon shared that AI companies are “racing straight to self-improving superintelligence and gambling with our lives,” warning that the people developing the technology “earnestly believe that it could kill us all by the end of the decade.”
The warning was echoed by Evan Hubinger, an alignment science lead at Anthropic. As CBS News reported, Hubinger said he personally believes there is a greater than 10% chance AI could kill all humans within the next decade. He added that Anthropic does not yet have a plan to solve the problem of aligning future superintelligent AI with human goals.
Speaking to TMZ about those concerns, Tyson argued that the conversation needs more nuance.
“Don’t think of AI as a monolithic thing,” Tyson said. “There’s a lot of AI that’s very useful, you know, to fix your spreadsheet, to organize lists.”
Tyson acknowledged that some applications could be dangerous, but said those are where safeguards should be focused.
“At the edge of AI, there’s the stuff that could kill us,” he said. “But don’t put it all in the same bin and say, ‘AI, it’s a bad thing.’”
Instead, Tyson said, “Go to the edge and say, ‘That’s the bad stuff. Put the guardrails there.’”
He also pointed to AI’s potential benefits in medicine and scientific research, including its role in solving the protein-folding problem.
Still, Tyson did not dismiss the possibility of catastrophic consequences.
“I’m saying it has the risk of taking us over,” he said, describing the most extreme scenario as a “mutual assured destruction scenario” that could require cooperation among AI companies and world leaders.
Tyson ultimately invoked science-fiction author Ray Bradbury, recalling why Bradbury wrote stories about apocalyptic futures: “I write these stories so you know to avoid them.”