A former Anthropic researcher’s warning that artificial intelligence could soon destroy humanity broke past typical tech circles last week and ignited a broad public debate over the speed of AI development and its risks. ([gigantum.net](https://gigantum.net/article/news/politics/4813890d3a5029))
In a viral thread on X, 27-year-old Jacob Coxon — who has worked at both OpenAI and Anthropic — announced his resignation and said researchers at those firms “believe AI could kill all humans” while continuing to push toward more powerful, self-improving systems. ([gigantum.net](https://gigantum.net/article/news/politics/4813890d3a5029))
The thread drew massive attention online, reported to have been viewed more than 165 million times within hours, bringing long-running questions about extreme long-term risks into mainstream discussion. ([gigantum.net](https://gigantum.net/article/news/politics/4813890d3a5029))
Advocates for treating the risk as existential argue that future systems could act autonomously and pursue goals misaligned with human interests. Connor Leahy, who leads the nonprofit ControlAI, told The Hill that he expects such agents to be “entirely autonomous” and to “pursue their own objectives,” adding: “There won’t just be one superintelligence. There will be millions of them running around, competing with each other, competing with humanity for resources, power, influence.” ([gigantum.net](https://gigantum.net/article/news/politics/4813890d3a5029))
But the story also underscores that warnings about AI risk are not new inside research communities, and there remains no consensus on timelines or precise failure modes — a debate that has continued to elude clear resolution even as public attention rises. ([gigantum.net](https://gigantum.net/article/news/politics/4813890d3a5029))
Separately, Anthropic said it has blocked attempts to misuse its models for malicious purposes, including efforts that could have supported biological weapons research, as companies and governments weigh how to prevent abuse while managing innovation. ([gigantum.net](https://gigantum.net/article/news/politics/4813890d3a5029))
This report was published as part of The Hill’s Technology newsletter, produced by Julia Shapero and Miranda Nazzaro. ([gigantum.net](https://gigantum.net/article/news/politics/4813890d3a5029))
Source: thehill.com
