Should AI pay for web content? Early proposal (HCCP)

2 aidanmorgan 1 8/24/2025, 8:17:57 AM github.com ↗

Comments (1)

aidanmorgan · 14h ago
AI systems are trained on and generate from vast amounts of web content—articles, code, images, datasets. These systems monetize at scale, but the creators whose work underpins them see no compensation. The concern is that without some mechanism of return, the open web risks becoming an unmaintained commons.

The Human Content Compensation Protocol (HCCP) is an early-stage proposal to address this. The principle is simple:

Human access remains free, content creators should monetise their work as they currently do.

Agentic AI systems that crawl, train on, or generate from content should be required to pay.

The draft outlines three layers:

1. Licensing — distinguishing human use from AI use.

2. Technical mechanisms — agent detection, request signing, decentralized identity.

3. Economic plumbing — micropayments, pooled settlements, bulk licensing.

This is very early, and that’s the point: protocols and standards only work if communities shape them. Like robots.txt or Creative Commons, the strength of HCCP will come from iteration, adoption, and feedback.

Request for support If you’re a developer, lawyer, economist, or creator:

Read and critique the proposal

Suggest improvements or alternatives

Contribute to the technical and legal design

Help test whether this can work in practice

The open web has always balanced openness with sustainability. The question is whether that balance will survive the AI era.