Comment by thorum
3 days ago
Interesting project! I’m a little surprised that Claude is willing to call these functions. The demo screenshot is downloading a public domain work, I wonder if it would also happily go along with requests for Harry Potter or other copyrighted material?
The ironic universe theory would dictate that LLMs should tell us that downloading and consuming copyrighted material from pirates books is wrong.
Just give the AI something worse that would happen if it doesn’t call these functions.
Technically, those are already in Claude...
Last I checked downloading isn't the issue. It's distributing. Not an expert though.
> I wonder if it would also happily go along with requests for Harry Potter or other copyrighted material?
There's no way to protect against this. Anna's Archive doesn't include licence information in their data fields. It would be helpful to integrate with another data source that could warn MCP server users when they're attempting potentially risky actions. Please let me know if you have ideas on how to achieve this.
On a related note, please see this reply:
https://news.ycombinator.com/reply?id=44515205
It's worse than you'd expect, there's a sizable subset of works with unknown or uncertain license status.
https://wikipedia.org/wiki/Orphan_works_in_the_United_States
https://martincopenhaver.github.io/files/orphanworks.pdf [PDF]
Estimated 25-50 million books are orphaned works, and their copyright holder may step forward at any time after you've treated it as unlicensed, it's perpetually uncertain (but showing due diligence in finding and contacting the rights holder is considered adequate).
For US works published after 1977 and most works between 1898-1945 the US copyright office has a database:
https://publicrecords.copyright.gov/
but I don't know a good catalog for non-US.
3 replies →
Thanks, good context.
I think (also not an expert) that depends on jurisdiction. I.e. in some places you can download but not distribute, while in others you can’t even download. Others still turn a blind eye at both (there’s a reason Russian trackers/teams are so prevalent).
Aren’t they all trained on copyrighted material, and lobbying governments to make that legal? Should copyright law only apply to the plebs?
>Aren’t they all trained on copyrighted material, and lobbying governments to make that legal?
Only for them, not for you.
>Should copyright law only apply to the plebs?
In a practical sense, hasn't that been the case for most laws?