Anthropic's Secret AI Cracked Every Operating System in Hours

Anthropic's Claude Mythos Preview — the most capable AI model ever
built — was quietly stress-tested by a security researcher at a
wedding in Bali. Within hours: zero-day vulnerabilities across
virtually every major operating system and browser on the planet.

Anthropic decided not to release it to the public.

Days later, Treasury Secretary Bessent and Fed Chair Powell summoned
the CEOs of the five largest US banks to an unannounced meeting at
Treasury HQ. JPMorgan. Goldman Sachs. Citigroup. Bank of America.
Morgan Stanley.

Only a handful of companies got access — under Project Glasswing.

Is AI moving faster than we can control it?

Sources: Bloomberg April 10 2026 · Bloomberg April 16 2026 · CNBC April 10 2026

#claudemythos #anthropic #ainews #artificialintelligence #aihistory

Видео Anthropic's Secret AI Cracked Every Operating System in Hours канала Nekai Global