# Why Data Collectors Must Pay As agents and LLMs reshape the web, a new contract between creators and machines must emerge. --- The rise of generative AI has sparked an explosion of data usage—and a silent crisis. Every day, bots and language models scrape, remix, and monetize content created by humans. Yet the vast majority of creators, educators, and developers receive nothing in return. This isn’t just an oversight. It’s a structural flaw. And it demands a structural solution. If machines use your work, they should also compensate you. Below, we unpack why data collectors must pay—from legal obligations to ethical responsibility to economic sustainability. --- ## Copyright Still Matters Scraping content doesn’t place it in the public domain. Nearly all online content is automatically protected by copyright. Using it—especially to train models or create derivative works—can violate intellectual property laws. Today’s legal landscape is evolving quickly: - OpenAI faces lawsuits from publishers, coders, and creators. - The EU’s AI Act and Digital Services Act demand training data transparency. - The U.S. Copyright Office is reviewing how AI-generated outputs relate to source works. Consent and licensing aren’t optional in the age of autonomous data collection. They are the legal foundation. --- ## This Is an Ethical Issue The current system rewards those who aggregate, not those who create. LLMs and agents generate billions of dollars by remixing blogs, papers, tutorials, datasets—and do so without attribution, permission, or payment. Ethical design demands more: - Creators should control how their work is used. - Usage should unlock value for the original author. - Recognition and compensation should be the default, not the exception. Without these principles, we erode the very trust and collaboration the open web was built on. --- ## The Economics Are Just as Clear AI systems rely on a constant stream of human-generated data. But if that content isn't sustainably monetized, the pipeline dries up. When creators are paid: - Quality rises - Participation grows - Agent-optimized content becomes viable We’ve seen this play out before: YouTube, Substack, and Twitch flourished by letting creators earn from attention. Now, we need the same evolution for agent-triggered interactions. We're excited to support and evangelize the adoption of `llms.txt`—a machine-readable metadata standard that enables content to declare licensing terms, usage rights, and tipping or compensation flows. It represents a critical step toward building monetization infrastructure for the agent-readable web. --- ## We’ve Done This Before The open web has a history of standardizing machine interaction: - `robots.txt` told bots where not to crawl. - API tokens and rate limits introduced access control. - AdSense turned passive content into income. What those formats did for Web2, `llms.txt` can do for Web3—and the emerging agent economy. It enables what some are calling **Generative Engine Optimization (GEO)**: content that is not only discoverable by AI, but legally and economically usable. --- ## Without Compensation, the Open Web Breaks If we don’t act: - Open contributors will stop sharing freely. - Content ecosystems will collapse into walled gardens. - LLMs will recycle increasingly outdated and low-quality data. - Trust between humans and machines will erode. The solution isn’t just policy or code—it’s an agreement. A machine-readable handshake between creators and the agents that use their work. --- ## Toward a Machine-Payable Internet Agents and LLMs are not going away. They are becoming first-class citizens of the web. The question is not whether they’ll use content—they already do. The question is whether they’ll do so fairly, legally, and sustainably. We believe the future of the web must include native support for compensation and attribution. That’s why we’re championing tools, standards, and infrastructure that make the internet machine-readable—and machine-payable. --- ***Want to integrate `llms.txt` into your publishing stack or agent tooling?*** Reach out to the OnChainSupply team and join the working group exploring the open standards for the machine-payable web.