Malware developers added nuclear and biological weapons text to to their spyware(twitter.com)
449 points by marc__1 3 days ago | 236 comments
tl;dr: Summary not available
HN Discussion:
  • LLM censorship around weapons is pointless since such information is already publicly accessible
  • AI guardrail triggers should be used as malware detection signals
  • This tactic warns that AI-based analysis tools need better sandboxing and security measures
  • ~Mocking the situation with jokes about poisoning endpoints or using magic refusal strings
  • Moderation/censorship mechanisms inherently double as denial-of-service vectors