Fears about artificial intelligence reached a new level after a former Anthropic researcher’s resignation post warned that AI companies could soon lose control of the systems they’re building. But not everyone is convinced an AI apocalypse is coming.
On Monday, President Donald Trump and Vice President JD Vance dismissed the sudden wave of industry alarm. Vance said the AI leaders’ calls for a slowdown felt like “a bit of a Trojan horse.” Trump went further, dismissing warnings that AI could take over the world and destroy humanity as a “HOAX, no different from RUSSIA, RUSSIA, RUSSIA” or his two impeachments. In an earlier post today, he argued the government already has enough regulatory power over AI firms and said China stands to benefit most from the current backlash against data centers.
Download the Straight Arrow app today to get the stories that matter free from manipulation, bias or agenda.™
Point phone camera here
But it isn’t only the politically interested who are skeptical. Researchers and industry figures have pushed back too — not by calling the claims fake, but by arguing they’re overblown given where the technology actually stands today.
Apocalyptic claims
On Sept. 8, Jacob Coxon, a 27-year-old researcher who spent three years in pretraining research at OpenAI and then Anthropic, announced his resignation on X. In his post, he warned that both companies were “racing straight to self-improving superintelligence and gambling with our lives.”
Coxon argued future systems could become capable of hacking critical infrastructure and acquiring real-world power. He said colleagues at both labs privately believe AI could kill everyone within the decade. The post has drawn well over 170 million views.
Hours later, Evan Hubinger, an alignment science lead at Anthropic, publicly backed Coxon. He said the odds of AI causing human extinction within a decade were above 10% and acknowledged the company doesn’t have a plan to align a superintelligent system. Hubinger also stressed that the currently deployed models pose very little risk to people.
Four days after Coxon’s resignation post, Anthropic CEO Dario Amodei published an essay calling for a slowdown. He proposed independent evaluators with insider-level access to frontier models and coordination among AI labs in democratic countries. OpenAI CEO Sam Altman echoed the call the same day, committing OpenAI to similar outside oversight.
Trump didn’t specify what “criminal and regulatory power” the government has over AI companies or how it addresses the concerns Amodei and Hubinger described.
Industry leaders also pushed back against the warnings. Nvidia CEO Jensen Huang called Coxon’s warnings “absurd” and “arrogant” at a Goldman Sachs conference on Thursday, according to Brad Gerstner, the founder of Altimeter, who attended the event. Huang said the labs have already completed more safety work than Coxon credits, but Nvidia is a major player in the AI space, making the most powerful chips for AI model training.
Jensen calls Jacob Coxon comments outlandish & “deeply untrue.” Said the labs are great, but his comments were wrong, arrogant, and ignorant of all the work being done around the industry to drive safety. 🧐🧐 pic.twitter.com/S253cHKQy7
— Brad Gerstner (@altcap) September 10, 2026
Santa Fe Institute computer scientist Melanie Mitchell told CBS News that AI technology wasn’t at the point where it could compete with humans in most ways. However, she said she was worried that AI companies weren’t aligning their models with goals that could best benefit people.
Others like cybersecurity researcher Artem Dinaburg argued that the real risk isn’t a runaway superintelligence but the potential for AI agents breaking into systems they shouldn’t be able to reach, something that has already happened multiple times.
While some in the AI space don’t believe Coxon’s claim was a hoax, they disagree on how seriously to take his concerns. For now, both Anthropic and OpenAI seem to be moving forward with more outside oversight as the government continues to debate more regulations for the technology.
