The US government and AI company Anthropic have been in a big fight for months. The Pentagon called Anthropic a security risk and banned its tools because the company wanted rules on how the military uses their AI, like no spying on Americans or killing without humans deciding. President Trump even ordered agencies to stop using them.
Then Anthropic launched Mythos, a super-smart AI that beats humans at hacking and finding software bugs. It can spot thousands of vulnerabilities in systems like operating systems and browsers, making cyber attacks easier and faster. Other AIs like GPT 5.5 are catching up quickly.
Now the White House wants Mythos for its own national security but blocks Anthropic from giving it to more private companies—from 50 to 120—due to compute shortages and risks. A new AI memo pushes agencies to use multiple AIs and eases some feud issues.
Tensions remain: Defense Secretary Pete Hegseth called Anthropic’s CEO an “ideological lunatic” who shouldn’t control military AI use. The White House met with Anthropic to calm things but walks a tightrope between access and a tough stance.
AI CYBER ARMS RACE: KEEP FRIENDS CLOSE, FOES CLOSER – OR RISK LOSING BOTH!
Sanjay Sahay
Have a nice evening.