Nicholas Carlini, a security researcher who long warned about the dangers of AI, is now leading defensive cyber research using Anthropic's frontier model "Claude Mythos 2 Preview" and arguing for limited distribution and a staged release. The reasoning: capabilities that would be dangerous in the hands of attackers should first be controlled by defenders.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.