← Back to the wire

The 'AI safety' movement is making AI less safe

SpeculationPolicySep 22, 2026

Anthropic alignment science lead Evan Hubinger publicly agreed that the company's product has a 10 percent chance of causing human extinction within a decade, following Jacob Coxon's resignation over similar fears. The article argues that "AI safety" advocates, shaped by effective altruism and the TESCREAL worldview Émile Torres describes, are pushing regulation that consolidates AI among frontier labs like Anthropic and OpenAI, echoing the AI 2027 scenarios, while making the technology less transparent and less safe.

Receipt № 20321 source · awaiting confirmation ◐

Evidence

1source· awaiting independent confirmation

No score is assigned. Sources and their independence are shown in the citation chain below.

Citation chain · 1 source

OpenAICompanyAnthropicCompanyJacob CoxonPersonEvan HubingerPersonÉmile TorresPerson
Canonical: https://reason.com/2026/09/22/the-ai-safety-movement-is-making-ai-less-safe/