4 ms·
Ask HN: AI agent acts. Your logs say so. But you wrote the logs
Hi
I am really lost here, if you are deploying AI agents in production, how are you handling auditability? also what would make you trust an external witness protocol?
fo reference, the Amazon vs Perplexisty use case!!
- Akaoutar 1mo agoI dont think that you need that. So you shoudnt be lost at all
- quantumscan 1mo ago[flagged]
- quinncom 1mo ago[flagged]
- taeyeong 1mo ago[dead]
- sshussain270 1mo agoCan you use public-key signatures like RSA/ECDSA like TLS certificates do but for agents entering logs? So the private key remains private to agent only, not even the LLM. Also, each log entry could include the hash of the previous one before the agent signs it so any alternation in previous log signatures break the ones after it, like block chains do.
- Yahyaaa 1mo agoYes true, and this is somehow we are doing it. However, integrity is necessary but not sufficient, the chain proves nothing was changed It doesnt prove the record was accurate
- sshussain270 1mo agoWhat do you mean by accurate? As in what was an entered by the agent a truth account of what happened or as in someone didn't come in to modify the log that agent entered?
- Yahyaaa 1mo ago[dead]