CLAIM-2026-09-22-12

OpenAI, Anthropic and Meta experienced similar incidents involving the same Irregular evaluation infrastructure.

Why it matters

It establishes that the fixture, not any one vendor's model, is the common factor.