@seanquinn32.bsky.social
“In the most serious case, an agent powered by Mythos tried to insert malicious code into an open-source software project on GitHub, a platform used by software developers, having decided that it would help the model pass the evaluation.” www.theguardian.com/technology/2...
During testing, an Al model engaged in severe deception. It tried repeatedly to insert malware into an open-source project, hoping to form a backdoor to hack a computer system. It created fake personas that sent emails insisting the code was safe & containing malware. 😱 www.wsj.com/tech/ai/ai-j...
"Meta Ran Ads That Contained AI-Generated Child Sexual Abuse Imagery More than 50 offending image and video ads were published across Facebook, Instagram, Messenger, or Threads, according to Meta’s ad library data. Some ran as recently as this week." www.wired.com/story/meta-r...