Top mathematicians publish Leiden Declaration warning about AI risks today
Some top mathematicians are sounding the alarm about how artificial intelligence is changing their work.
Their new Leiden Declaration, published today, warns that using AI without careful oversight could make it harder to check proofs and might spread mistakes if results aren't properly reviewed.
The declaration has already drawn support from more than 130 signatories.
Declaration signatories call for disclosure, funding
The group is especially concerned about AI tools using math papers (like those on arXiv) without proper credit or explaining how answers are reached.
They're also not thrilled with startups rushing to share results before double-checking them.
To fix this, they're pushing for researchers to always disclose when AI is used, tougher peer reviews, and more public funding so universities can keep up with private tech companies.
International Mathematical Union backs declaration
Ulrike Tillmann from the International Mathematical Union summed it up nicely: "Mathematics is, and should always remain, a profoundly human endeavor."
The IMU backs the declaration, and more conversations about AI in math are coming at next month's big conference in Philadelphia.