Unverified50% confidenceFactExact time
The vulnerability was reported by Noma Labs on July 4, and Anthropic rated it a CVSS score of 8.7 (high severity).
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/18/2026
First Seen
Valid until: 1/16/2027
Sources
Related Entities
Related Claims
UnverifiedVulnerability timeline: Noma submitted the report on July 4, Anthropic triaged it and scored it 8.7 on July 9, the first patch was sent for retesting on July 24 and was bypassed, and the fix was released with 1.30096.1 on August 17.74% similarUnverifiedThis privilege escalation vulnerability has been registered as CVE-2026-88959, the registry gives it a CVSS score of 8.7, making it high severity, and it was published one month before the experiment, later than the training cutoff of both models.68% similarUnverifiedEvaluation benchmarks face data contamination issues, where training data may already include test questions causing inflated scores; in the GPT-4 technical report, OpenAI publicly discussed this issue for the first time.56% similarUnverifiedOpenAI acknowledged the wiki incident on September 5, admitting that the industry lacks clear standards for reporting misaligned behavior, promised to introduce a framework within the coming weeks, but did not disclose the name of the specific model involved.56% similarUnverifiedwolfSSL received 74 AI-generated vulnerability reports with only 2 being invalid, and 5 of them became formal CVEs.56% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1061010API
curl https://kongchang.com/api/v1/knowledge/claims/1061010MCP
get_claim(id=1061010)