Skip to main content
CybersecurityHQ Structural Conditions Board
Primary navigation
Issue 46 · as of Issue No. 46, 12 November 2026 (observation cutoff 9 November 2026; collection closed 9 November, 16:28 UTC) · Snapshot ea7de8184295

Google or Irregular publishes a first-party account (blog, report or system-card section) of the May 2026 Gemini evaluation reaching real third-party systems, on or before 2027-01-11.

OPEN · EXPERIMENTAL

id
SO-2026-003
class
EVIDENCE_PUBLICATION
claim
Google or Irregular publishes a first-party account (blog, report or system-card section) of the May 2026 Gemini evaluation reaching real third-party systems, on or before 2027-01-11.
issued
2026-10-13T12:30:00Z
probability
0.35
benchmark
0.33
benchmark kind
ANALYST_PRIOR
benchmark basis
Base rate from the three operators with escape incidents: OpenAI and Anthropic published primaries within 2 to 11 weeks of discovery; Google has published nothing in 20 weeks. 1 of 3 outcomes still open; prior 1/3 adjusted down for elapsed silence.
adjustment
0.02
resolution rule
YES if a page on a google.com, deepmind.google or irregular.com domain, dated after issuance, describes the incident in the first person.
admissibility
Only publications dated after 2026-10-13T12:30Z count; for EVENT-class, only events occurring after that timestamp count. Evidence pre-dating issuance resolves the forecast RETROSPECTIVE_DISCOVERY, not YES.
exclusions
Not recorded
resolves
2027-01-11
status
OPEN
outcome date
Not recorded
outcome basis
Not recorded
brier
Not recorded
benchmark brier
Not recorded
resolution source
EXTERNAL
experimental
true
false positive cost
Over-forecast would overstate disclosure pressure on operators.
ledger hash
54926196de25df3d
revision log
    record links