Posted in

AI ‘that could escape the lab’ sparks fear in the City

Anthropic’s Mythos is so adept at cyber hacking that it is too dangerous to release to the public

The accidental leak of news about an all-powerful hacking bot will have done little to calm the fears of AI doom-mongers.

Days before the official announcement of Anthropic’s Claude Mythos, reporters from Fortune magazine uncovered an online datastore of unprotected files belonging to the tech company, including a draft blog revealing the existence of the new AI.

Since the botched reveal earlier this month – blamed on “human error” – industry leaders and government experts have been sounding the alarm over the powerful bot.

Our next nightmare. Worse may be just a couple of years down the road as AI makes breakthroughs in Quantum computing.

Share