Future of Privacy Forum launches roundtable series on privacy risks of agentic AI
- Severity
- Info
- Sources
- 1
What happened
No vulnerability or CVE is involved; this is an event report on privacy risk in agentic AI. A draft taxonomy of 24 agentic privacy risks across five domains was released as version 0.1 for public feedback, intended to support pre-deployment benchmarks. Speakers flagged consent fatigue, the lack of standard protocols for agents to negotiate privacy terms, deletion requests not reaching downstream agents, and weak agent identity and human-approval verification. Proposed controls include cryptographically binding agents to authorising users, machine-readable intent, deterministic enforcement and zero-knowledge proofs, with data minimisation applied first.
For security analysts · Written by AI from 1 source · Wrong?
What to do
- No action needed.
Written by AI from the sources, checked against them
Full story timeline · 1 update
Who reported · 1 outlet
Who and what is in it
Government agencies