alt Hacker News> certainly it would be very difficult to prove liability on OpenAI for this
My understanding is that OpenAI products specifically provided help in planning attacks / self harm.
> certainly it would be very difficult to prove liability on OpenAI for this
My understanding is that OpenAI products specifically provided help in planning attacks / self harm.
Full transcripts are unfortunately not available for any of those cases, but from what I've found it provided general information about e.g. how to load and operate a firearm or how past mass shootings have been received in the media.
The way I see it, providing general information is not a crime. They're basically saying: "Oh no! My repository of all human knowledge contains all human knowledge! It must be defective!"