# Former Anthropic Researcher Warns AI Could Threaten Humanity > Source: > Published: 2026-09-10 13:27:54+00:00 **September 10, 2026**, (Inside AI) — A former Anthropic researcher has publicly accused the AI lab and its rivals of acting irresponsibly, warning that the technology could threaten humanity. The claim arrives during a week dominated by geopolitical and economic turmoil, raising the stakes for AI governance debates. The researcher, who resigned from Anthropic, made the remarks on a podcast episode that also covered the Iran war timeline and fresh inflation concerns. The episode aired as U.S. President Donald Trump proposed $5,000 checks for every American, a move that could pressure bond markets. The European Central Bank is expected to raise interest rates again as energy prices climb. That macroeconomic backdrop sharpens the contrast with the AI researcher's warning: immediate financial pain versus a slower-burning existential risk. Anthropic has not publicly responded to the researcher's specific allegations. The company has previously emphasized safety research and a cautious approach to deploying powerful models. Rival labs, including OpenAI and Google DeepMind, have faced similar criticism from former employees over the past three years. The researcher's core claim is that current AI development outpaces safety measures. That argument echoes warnings from other prominent figures, including Geoffrey Hinton and Yoshua Bengio, who have called for stronger oversight of frontier AI systems. Industry critics often push back on such warnings. They argue that focusing on hypothetical existential risks distracts from real harms like bias, misinformation, and labor displacement. The debate has split the AI research community since at least 2023. The podcast episode did not name the researcher. It also did not specify which Anthropic project or model triggered the resignation. That lack of detail leaves room for competing interpretations about the severity of the warning. Anthropic's corporate structure differs from some rivals. It operates as a public benefit corporation, a legal form designed to balance profit with social responsibility. That structure has not shielded it from internal dissent. The broader AI safety conversation has shifted since 2025. Several governments have introduced mandatory testing requirements for large models. The European Union's AI Act now imposes binding rules on high-risk systems, though enforcement remains uneven. Trump's $5,000 check proposal could inject new volatility into tech valuations. AI companies depend on cheap capital for expensive training runs. Higher inflation and bond yields would make that capital harder to secure. The European Central Bank's expected rate hike adds another layer of pressure. European AI startups already face higher energy costs for data centers. Tighter monetary policy could slow their expansion plans. Despite these headwinds, frontier AI labs continue to raise billions. Investors appear willing to fund ambitious projects even as safety concerns mount. That disconnect is exactly what the former Anthropic researcher flagged. The podcast also referenced a column on dollar policy, linking currency swings to tech sector performance. A strong dollar typically hurts U.S. exporters but helps import-heavy AI supply chains. The relationship is rarely straightforward. Listeners seeking deeper economic context can follow the [Reuters markets coverage](https://www.reuters.com/markets/) for ongoing analysis of inflation and central bank moves. The episode's production credits list Eliza Davis Beard and Abisoye Osundairo as producers. Sound engineering was handled by Sebastian and Josh Sommer. The podcast is available on Spotify and the Reuters app. The AI researcher's warning may not change corporate behavior overnight. But it adds another voice to a growing chorus demanding more transparency from frontier labs. Whether regulators act on that pressure remains an open question.