The data on which authors are part of which arxiv papers is already in the arXiv database, and in Google Scholar, and in other libraries. I appreciate that it's not an easy task to get that as a third party. But the burden should be on the operators of alphaxiv to figure out a solution for this platform to take off, not for me as a user?

Yea I agree it shouldn't be on you as a customer, was more asking out of curiosity.

I don't think Google scholar has this fully solved either, I've seen many misattributed papers there.

The only way I see this working is for paper authors to include their public keys in the paper; preferably as metadata and have them produce a signed message using their private key which allows them to claim the paper.

While the grandparent is understandably disappointed with the current implementation, relying on emails was always doomed from the start.

Given that the paper would have be changed regardless, including the full email address is a relatively easy solution. ORCID is probably easier than requiring public keys and a lot of journals already require them.

W3D Decentralized Identifiers are designed for this use case.

Decentralized identifier: https://en.wikipedia.org/wiki/Decentralized_identifier

W3C TR did-core: "DID Decentralized Identifiers 1.0": https://www.w3.org/TR/did-core/

W3C TR did-use-cases: "Use Cases and Requirements for Decentralized Identifiers" https://www.w3.org/TR/did-use-cases/

"Email addresses are not good 'permanent' identifiers for accounts" (2024) https://news.ycombinator.com/item?id=38823817#38831952

I'm sure that would work, but most researchers already have an ORCID are required to provide it other places anyway.

In comparison to DIDs, ORCIDs aren't sk/pk pairs that can be used to cryptographically sign.

A person can generate (and optionally register) additional DIDs if they please.

A person can request additional ORCIDs if they please

Probably it would be better to merge the project with Arxiv and tie it to the accounts there.

The claiming was 'solved' and ORCID, which both basically do no checking at all. It's just a yes/no clicking for fuzzy matched author name lists. So I guess it is enough until there is a dispute. If you are important enough to be the target of trolls than you are in a league beyond most research platforms.

Perhaps by linking to their actual arxiv id?

There should be an equivalent of S/MIME for researchers if e-mail is not accessible.