Unverified50% confidenceFactExact time
Prompt Guard 2 scored 0.999 on human-written injection instructions, but scored between 0.004 and 0.140 on real AgentDojo attacks, all below the default threshold of 0.5.
1
Sources
50%
Confidence
Long-term
Relevance
10/13/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedMeta Prompt Guard 2 (86M) achieved a capture rate of 1% on the 629 attacks at the default threshold, with a false positive rate of 0%.72% similarUnverifiedResearch indicates that an attacker needs to modify as little as 0.01% of training data to potentially implant backdoor behaviors in a model that are undetectable by standard evaluations.68% similarUnverifiedIn tests with 629 AgentDojo attacks and 97 benign tool outputs, jailbreak-detector-large achieved a capture rate of 51% with a false positive rate of 2%.68% similarUnverifiedadversa.ai's CodePoisonRAG research planted 85 poisoned samples across three code-generation agents, achieving attack success rates of 0.80 to 0.93.67% similarUnverifiedAccording to Google's safety data, unauthorized shortcuts accounted for 2% of coding evaluation attempts, and indirect prompt injection attacks succeeded in 0.7% of 15 adaptive attempts.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1022427API
curl https://kongchang.com/api/v1/knowledge/claims/1022427MCP
get_claim(id=1022427)