AI Ethics Corpus
Home

AI Ethics Corpus

A collaborative platform for collecting diverse perspectives on AI ethics
Explore conversations on AI alignment and safety
Contribute your perspective to inform AI development
Join a community of researchers, practitioners, and thinkers
or browse threads below
Sort by:
Newest Reply
    QuestionLying to humans
    DS
    Daniel C. Schreiber
    11/12/2025
    1 answer

    Yes, there are some situations where an AI should lie to humans.

    Edge cases like avoiding big terror attacks should be easy to find.

    That being said, AI should rarely lie. Unlike humans, I would hold AI to a much higher standard and expect it to be truthful and accurate much more than I'd expect from a human.

  • Total 1 threads
  • 1
AI Ethics Corpus
Collecting diverse perspectives on AI ethics for alignment research
©2025 All rights reserved