The International Mathematical Union released the Leiden Declaration on June 2, 2026, warning that unchecked artificial intelligence integration threatens the integrity, autonomy, and ethical standards of mathematical research.

Key Points

• The declaration warns that AI models generate plausible but incorrect proofs, potentially cluttering scientific literature with flawed results that are difficult to verify.
• Signatories highlight that AI training often violates copyright protections and fails to properly credit human researchers for their original work.
• The document cautions that tech companies are gaining excessive influence over research priorities, which may shift focus toward AI-amenable problems rather than fundamental mathematical inquiry.
• Researchers are urged to maintain accountability for their work, disclose AI usage, and resist pressure to collaborate with corporations on asymmetric terms.
• The declaration calls on policymakers to regulate the AI industry, protect author rights, and invest in public computational infrastructure to reduce reliance on private tech firms.

Why it Matters

This declaration signals a growing rift between the academic mathematical community and the rapid, often opaque, deployment of commercial AI tools. By establishing these guidelines, mathematicians aim to preserve the discipline's rigorous standards of proof and prevent corporate interests from compromising the long-term future of scientific research.