5 takes tagged with this topic.
Anthropic's Mythos model was supposed to be so dangerous at finding security vulnerabilities that they couldn't release it to the public. That was the headline in April.
This morning I posted about indirect prompt injection. Attackers hiding invisible instructions in web pages to hijack AI agents.
McKinsey built an internal AI platform called Lilli. Three quarters of their 40,000 employees use it for strategy work, client research and document analysis.
The web is being weaponized against your AI agents. And the attack is hiding in plain sight.
An AI coding agent just deleted an entire company's production database and all its backups. In 9 seconds.
