Skip to content
InfoNews

Do AI’s mathematical advances put PKI under threat?

Published
Record updated
View JSON

Summary

Johns Hopkins cryptographer Matthew Green warned on X that public-key cryptography could fail because AI is beginning to outperform humans on some mathematical problems. He estimated a 15% chance that confidence in existing public-key encryption algorithms is functionally lost, and he flagged that most standardized public-key encryption rests on a small number of hard problems. Tampere University cryptographer Markku-Juhani O. Saarinen takes the warning seriously but urges caution against hasty replacement of established standards.