White House worried about compute limits as it blocks wider access to Anthropic's Mythos

The White House rejected Anthropic’s bid to expand access to its Mythos AI model to about 70 additional companies.

Access remains limited to roughly 50 organizations as officials weigh security concerns, compute shortages, and ongoing government use.

The move underscores the tension between wider AI deployment and concerns over model security, infrastructure capacity, and supply-chain risk.