According to multiple media reports, Amazon CEO Andy Jassy raised concerns about Anthropic's Mythos-class AI model with senior White House administration officials on Thursday, June 11, 2026. The warning came after Amazon researchers used a series of prompts that successfully elicited information about cyberattacks from the model.
The reports, which cite unnamed sources familiar with the matter, indicate that Jassy's direct intervention led to the White House taking action to shut down the model. Anthropic, an AI safety company, had been developing the Mythos model as part of its next-generation AI systems.
Neither Amazon nor Anthropic have officially commented on the reports. The White House has not confirmed the meeting or any subsequent actions. The incident highlights ongoing tensions between AI developers and safety researchers over the potential misuse of advanced language models.
