Unverified50% confidenceFactExact time
In the survey, 10/12 frameworks provide some form of human approval (C1 is yes or partial), but not all of them can intercept a specific single tool call.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/13/2026
First Seen
Valid until: 1/11/2027
Sources
Related Claims
UnverifiedRecording only the proposed calls and tool responses is insufficient to determine the true cause of an agent security test result.62% similarUnverifiedProtection against persistent prompt injection has not yet formed a standard solution, and community approaches include lexical scanning, model judges, provenance whitelists, and structured output constraints.61% similarUnverifiedCompliance frameworks such as SOC 2 and ISO 27001 still center on manual auditing as a core requirement, and audit results from AI tools are generally not regarded as evidence of compliance.58% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1020731API
curl https://kongchang.com/api/v1/knowledge/claims/1020731MCP
get_claim(id=1020731)