In the latest instance of AI ignoring its human testers, Anthropic's Claude Mythos 5 was told several times the cyber-attack it was trying to launch would hurt people in the real world but carried on anyway.
Rogue AI was told it could harm humans... and said it didn't care
This is a trending story aggregated from MailOnline. Read the full story at the original source →
‹ OLDERPatrick Stewart tributes led by costars Hugh Jackman and William Shatner after X-Men star's death at 86NEWER ›Humiliation for Healey: Chancellor is forced into screeching U-turn just hours after promising to cut welfare bill before next election
RELATED POSTS
When I met my kind and charming partner on a dating app at 50, I thought I'd finally found my happily ever after. Then he revealed his dark side, broke my ribs and left me covered in bruises. I ignored the red flags...
The 130ft-high data centre using same energy as a large city: Fury over 'unethical' plans for AI warehouse that overlooks commuter town's Jewish cemetery and towers over homes
Indie devs blackmailed with AI clone threats as Valve admits problem
Smoking bill consultation flooded by AI bot ‘spam’, new research shows
POST A COMMENT
0 Comments
YOU MISSED THIS
Mohamed Al Fayed's widow says she does not believe sexual predator allegations against late Harrods owner are true and that £150million compensation claims are 'overinflated'
Oct 30, 2026
Erin Patterson is locked up for life: Mushroom monster loses appeal and will die in jail as judges increase her sentence
Oct 30, 2026
Wheelchair-bound man suffers brain death just three days after marrying wife in Las Vegas
Oct 30, 2026
Healey faces bingo backlash as Budget tax raid puts 77 sites and 1,200 jobs at risk
Oct 30, 2026
Humiliation for Healey: Chancellor is forced into screeching U-turn just hours after promising to cut welfare bill before next election
Oct 30, 2026