X XFediYour place in the Fediverse
XFEDI

Conversation

ActivityPub

A weak key is an event. A weak feedback loop is a condition.

Eight eMudhra CA keys could be recovered by trying fewer than 256 values each. (A good P-384 key has ~10^115 possibilities.) The algorithm was fine; generating or selecting the key wasn't.

Across seven public Bugzilla reports (different severities, no proven common cause), one question recurs: was the rule enforced, verified, and tested after the fix?

Timeline and sources: https://banbreach.com/insights/emudhra-certificate-incidents/?utm_source=mastodon&utm_medium=organic_social&utm_campaign=controls_on_trial_01&utm_content=thread_pattern #InfoSec #PKI #Java

BanbreacheMudhra CA Key Compromise: Timeline, Java Impact, Lessons | BanbreachSeven Mozilla bug reports on eMudhra, eight recoverable CA keys and Oracle’s removal of its roots from Java. A sourced timeline and what to ask providers.
↩ Reply ♡ 0 · ↻ 0 ↗ Original

Oct 26, 2026 · 08:15 UTC

Replies
↳ In reply to a post

Practical angle for platform teams: on 24 Oct, Oracle's JDK update removed eMudhra's G1, G2 and G3 roots from the default Java truststore. A precaution, not proof that all three roots were compromised, but it's how a CA incident becomes an application incident. Do you know which of your services trust which roots?

#Java #PKI

↩ Reply ♡ 0 · ↻ 0 ↗ Original