qwant news | White House fights Anthropic's plan to expand Mythos tool that experts fear could cause AI doomsday
AI generated summary, Read the full article for complete information.
The White House is pushing back against Anthropic’s proposal to widen access to its powerful Claude Mythos AI tool, warning that expanding the rollout from the current handful of government users to an additional 70 companies—bringing the total to 120 under “Project Glasswing” with firms such as Amazon, Google and JPMorgan—poses serious security risks. Anthropic itself has warned that Mythos could be used to hack electric grids, power plants and hospitals, and recent reports say a few users have already breached the system. Trump‑administration officials also doubt Anthropic has enough computing capacity to serve both the government and the new corporate clients, and the White House is wary after the Pentagon terminated Anthropic’s contract earlier this year over concerns about unrestricted AI use. Meetings between Anthropic CEO Dario Amodei and White House officials aim to balance innovation with safety, while the company continues to grant early access to security researchers and strike deals with major tech firms to increase its compute resources.
Read more:
https://nypost.com/2026/04/30/business/white-house-fights-anthropics-plan-to-expand-mythos-tool-that-experts-fear-could-cause-ai-doomsday/
#WhiteHouse #Anthropic #Pentagon #DarioAmodei #ClaudeMythos #SusieWiles
