---
title: "Discussion on \"Why 'monitoring' isn't enough for AI agents — and how I made delegation cryptographically verifiable\" | Hashnode"
description: "Discussion on \"Why 'monitoring' isn't enough for AI agents — and how I made delegation cryptographically verifiable\". As AI agents start delegating tasks to each other, logging what they do isn't proof of who authorized what. Here's how I used Ed25519 signatures to make agent delegation verifiable in the browser — no"
url: "https://hashnode.com/posts/why-monitoring-isn-t-enough-for-ai-agents-and-how-i-made-delegation-cryptographically-verifiable/6ab39a033cef54781e67047c"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [Why 'monitoring' isn't enough for AI agents — and how I made delegation cryptographically verifiable](https://provenza.hashnode.dev/why-monitoring-isn-t-enough-for-ai-agents-and-how-i-made-delegation-cryptographically-verifiable?utm_source=hashnode&utm_medium=feed)

As AI agents start delegating tasks to each other, logging what they do isn't proof of who authorized what. Here's how I used Ed25519 signatures to make agent delegation verifiable in the browser — no

Sep 23 · 6 min read · provenza.hashnode.dev  
Read the article

-   [#security](https://hashnode.com/tag/security)
-   [#ai](https://hashnode.com/tag/ai)
-   [#governance](https://hashnode.com/tag/governance)
-   [#opensource](https://hashnode.com/tag/opensource)
-   [#pii](https://hashnode.com/tag/pii)

## Written by

[L'ka](https://hashnode.com/@provenza)

@provenza

  
[View profile](https://hashnode.com/@provenza)

## Responses

Markdown supported · ⌘↵ to post

No responses yet.