--- title: "The Mind-Weapon Panic: From CIA Psychics and UFOs to AI Apocalypse" url: "https://usubjects.com/article/the-mind-weapon-panic-from-cia-psychics-and-ufos-to-ai-apocalypse-4320" updated: "2026-10-22T16:23:16.671Z" description: "The article examines recurring fears about “mind weapons,” linking historical claims involving CIA psychics and UFOs with contemporary concerns about artificial intelligence. The provided material contains no further details about its arguments or evidence." sources: ["https://www.skeptic.com/article/the-mind-weapon-panic-from-cia-psychics-and-ufos-to-ai-apocalypse/"] licence: "All rights reserved" --- # The Mind-Weapon Panic: From CIA Psychics and UFOs to AI Apocalypse Published: 2026-10-21 Source: [Google News](https://www.skeptic.com/article/the-mind-weapon-panic-from-cia-psychics-and-ufos-to-ai-apocalypse/) ### A New Warning in an Old Pattern A public warning from a former artificial-intelligence researcher has revived a familiar question: **how should the public distinguish a genuine security threat from a story amplified by fear, institutional interests and limited evidence?** In September 2026, 27-year-old Jacob Coxon resigned from Anthropic and posted a series of warnings on 𝕏, arguing that some of the people developing the most advanced AI systems sincerely believe the technology could kill humanity before the end of the decade. His comments attracted more than 115 million views by the time Axios reported on his departure. Coxon emphasized that he had left two months before his equity was scheduled to vest. “I left before any of my equity vested,” he told Axios. “I no longer have anything to gain by juicing up Anthropic’s valuation.” Coxon’s warning was reinforced by other figures with direct experience in AI laboratories. Evan Hubinger, who leads Anthropic’s Alignment Science team, wrote, “We really do earnestly believe AI could kill all humans,” adding that he personally considered the probability to be greater than 10 percent within the next decade. Bilal Chughtai, a former Google DeepMind safety engineer, resigned in July 2026 after expressing similar concerns, while former Anthropic safety team member Joe Benton also raised alarms. ### From Paranormal Claims to AI Catastrophe The broader argument places these warnings within a longer history of **“mind-weapon” panics**—fears that hidden technologies, secret psychological programs or unexplained phenomena could fundamentally alter human behavior or threaten civilization. The article connects contemporary AI anxiety with earlier claims involving CIA psychics, UFO research and state-sponsored paranormal investigations. That comparison does not establish that the technologies or claims are equivalent. Instead, it focuses on the information environments surrounding them. During the Cold War, investigations of unusual observations existed alongside classified aerospace programs, deliberate disinformation and genuine scientific inquiry. As the article observes, “Military secrecy generates mythology that sometimes provides useful cover for military secrecy and increased defense spending.” The coexistence of authentic evidence and fabricated narratives can make it difficult for outsiders to determine what is known, what is speculative and what may be strategically obscured. ### Evidence, Institutions and Incentives The same problem now appears in debates over AI safety. Some members of Congress responded to the resignations by calling for new legal safeguards, while others questioned the warnings and the institutions promoting them. Critics have highlighted financial and organizational links involving Anthropic, its investors, journalists and independent evaluators. Dustin Moskovitz, an early Anthropic investor, serves on the board of Coefficient Giving, formerly known as Open Philanthropy. The organization funds the Tarbell Centre for AI Journalism and METR, an independent evaluator that has worked with Anthropic. **Those connections may invite scrutiny, but funding relationships alone do not prove that journalists or evaluators have abandoned professional independence.** Anthropic has also donated $40 million through two contributions to the super PAC Public First Action, which supports stronger federal AI regulation. The company’s July 2026 lobbying disclosure reported $2 million in spending over three months, $400,000 more than in its April filing. Former U.S. diplomat James Andrew Lewis argued that regulation could benefit established companies by making it harder for competitors to enter the market. “He might actually believe some of this stuff,” Lewis said of Anthropic co-founder Dario Amodei, “but I’d say self-interest really explains it more.” ### The Unresolved Question The dispute is therefore not simply whether AI could pose an extreme risk. It is also about **how warnings gain authority**, how institutions respond to them and whether commercial or political incentives shape the narrative. The history of UFOs and paranormal programs suggests that secrecy can conceal both important facts and deliberate mythmaking. For the public, the central challenge remains separating evidence from interpretation. AI researchers may sincerely believe catastrophic outcomes are possible, and their technical experience makes those warnings consequential. At the same time, institutional funding, regulatory strategy and public fear can reinforce one another. The article’s broader conclusion is cautionary: in an increasingly fragmented information environment, neither a dramatic warning nor the motives of its messenger should be accepted without examining the underlying evidence. ## Provenance This USUBJECTS page is a summary or digest of the linked original source. Any accompanying generated visual is an illustrative reconstruction, not evidence.