OpenAI and Anthropic share findings from a joint safety evaluation

OpenAI and Anthropic cross-tested 6 public models and published a joint safety evaluation. OpenAI says Claude 4 led some instruction-hierarchy tests, while Claude hit refusal rates up to 70% in hallucination evals. Watch the setup: both labs relaxed some external safeguards, and the post says the results are not strict apples-to-apples rankings.

Why it matters: HKR-H/K/R all pass: rival frontier labs jointly evaluating six public models is inherently clickable, and the post adds five test categories plus a 70% refusal datapoint. This is a strong safety research release, not a model launch or executive move, so it lands in featured, notp