Unverified claim: Anthropic has cracked popular encryption algorithms

zephyr_z9 · x · 2026-11-20

User @zephyrz9 claimed on X that Anthropic has internally broken several popular encryption algorithms but will never release it due to the chaos it would cause; the real worry, they say, is open models gaining the same capability next year. The quoted @CXCarroll responded that if true it would be wild, though it wouldn't literally end cryptography — SHA-256, for instance, doesn't rely on prime number theory at all.

Note: this is an unverified personal claim with no supporting evidence.

Related event: Anonymous Claim: Anthropic Has Broken Major Encryption Algorithms(3 posts)→

Original post →

More from Models

Models channel →