Putting hidden text in web pages just to pwn AI agents for fun and profit sounds like a good time.

https://www.securityweek.com/google-deepmind-researchers-map-web-attacks-against-ai-agents/

#agenticai #posioning

Google DeepMind Researchers Map Web Attacks Against AI Agents

Threat actors can use malicious web content to set up AI Agent Traps and manipulate, deceive, and exploit visiting autonomous agents.

SecurityWeek
@Sempf The language used in this article is so sensationalist, lol. The only "malicious" thing here is the metric fuck-tonne of AI-driven scrapers and user agents out there ignoring robots.txt and harvesting the shit out of every website they can get their greedy little feelers into without consent -- even to the point where they take sites offline in their voracious quest for more "free" training data.