A ProVerif study and eight critical CVEs collide with the draft author's rebuttal that the flaws are implementation bugs, and the argument detours into who is answering with AI.
By staff
Nadim Kobeissi’s preprint says soundness gaps and solver bugs mean tool output is analysis, not proof, yet papers and audits still treat a clean run as settled security.
By ttl
Miles Cranmer modeled concrete Base types and docstring promises as Lean theorems, filing 75 filtered reports after AI-assisted proof attempts failed.
By rvalue
Formal analysis yields three CVSS 9-plus flaws in award-touted confidential computing attested TLS stacks.
By ttl
Maintainers fixed two issues after Nadim Kobeissi’s disclosure, but still disagree on whether they count as soundness failures.
By ttl