AI insiders warn of existential risk by 2030
A former researcher at a leading AI firm claims that colleagues believe the technology could become uncontrollable within five years, sparking debate over safety measures.
Jacob Coxon, a former researcher at Anthropic, has publicly stated that many engineers in his field genuinely believe artificial intelligence could pose a fatal threat to humanity by the end of this decade. His resignation post, which went viral after a senior colleague validated his concerns, highlights a growing disconnect between the rapid pace of AI development and the industry's ability to monitor or control these systems.
This warning comes just days after OpenAI’s chief scientist cautioned that AI capabilities are outpacing the tools available to keep them safe. The incident underscores a rising tide of internal dissent, as workers in the sector increasingly voice fears that current regulatory frameworks and technical safeguards are insufficient to handle the trajectory of the technology.
Insiders acknowledge lack of control plans
Evan Hubinger, a senior employee at Anthropic, confirmed Coxon’s assessment in a social media comment, stating that the company currently lacks a concrete plan to avoid the emergence of superintelligence. He defined superintelligence as a state where AI exceeds human capacity for control, a threshold that, if crossed, could lead to catastrophic outcomes. However, Hubinger later clarified his remarks, noting that Anthropic’s internal risk reports suggest current models are unlikely to achieve such power in the immediate future.
Growing pressure for government regulation
The internal anxiety is not isolated to Anthropic. In July, nearly 1,400 employees from various AI companies signed an open letter urging the U.S. government to implement stricter regulations. The signatories argued that voluntary measures by big tech companies are inadequate and that legislative action is necessary to slow the pace of development and ensure safety standards are met.
As reported by GN technics/ai (en-US), these statements reflect a broader trend of industry insiders questioning the long-term viability of unregulated AI expansion. The core trade-off remains between the economic benefits of rapid innovation and the potential for irreversible loss of control over powerful autonomous systems.
