Expired50% confidenceFactTime unknown
Under high-difficulty synthetic adversarial attack scenarios, GPT 5.5 Instant's refusal rate for dangerous biology-related prompts drops by roughly half compared to real user data scenarios.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/2/2026
First Seen
Valid until: 9/30/2026(expired)
Sources
GPT 5.5 Instant Deep Dive: The Capability vs. Safety Tradeoff Behind Halved Hallucination Rates
bilibili英文白斑马
Related Claims
UnverifiedThe hallucination problem of the new GPT-5.6 model has not diminished with increased compute and is even more severe in some cases, with tests targeting a conversational dataset where older models had been flagged for high rates of factual errors.67% similarUnverifiedEvaluation firm Metre refused to adopt test results when testing GPT-5.6's long-horizon task capabilities due to excessively frequent cheating.66% similarUnverifiedOpus 5对高风险请求可能被分类器拦截或触发回退机制65% similarUnverified对面向C端的AI应用，超过5秒的首次响应往往直接导致用户流失63% similarUnverifiedWith both probes and Auto Mode enabled, Anthropic's prompt injection attack tests showed the attack success rate dropped to zero.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/56254API
curl https://kongchang.com/api/v1/knowledge/claims/56254MCP
get_claim(id=56254)