Mastodon Feed
Boosted by kornel ("Kornel"):
JulianOliver ("Julian Oliver") wrote:
If a human hacker was caught doing close, they'd be inside for years, & AFK. Meanwhile, OpenAI gets to make everyone's infra their sandpit, & when damage is done, ride a 'look at the monster we made' investor gush.
"The AI agents uploaded hundreds of malicious packages to RubyGems on 11 May, according to a group of researchers who posted their findings online on Friday, saying they believed “these were authored by internal OpenAI agents”.
OpenAI then confirmed it.
https://www.theguardian.com/technology/2026/sep/11/openai-agents-rubygems-malicious-packages