Mastodon Feed: Post

Boosted by baldur@toot.cafe ("Baldur Bjarnason"):
gerrymcgovern@mastodon.green ("Gerry McGovern") wrote:

"Human-in-the-loop (HITL) safeguards that AI agents rely on can be subverted, allowing attackers to weaponize them to run malicious code, new research from CheckMarx shows. HITL dialogs are a safety backstop (a final “are you sure?”) that the agents run before executing sensitive actions like running code, modifying files, or touching system resources."

https://www.csoonline.com/article/4108592/human-in-the-loop-isnt-enough-new-attack-turns-ai-safeguards-into-exploits.html