Content Signals
Open framework for declaring content usage preferences to AI systems via HTTP headers or meta tags (ai-train, search, ai-input).
Definition
Content Signals is an open framework (announced by Cloudflare in September 2025) that allows content publishers to express how their content may be used after it has been accessed by an AI system. It addresses a gap left by robots.txt: robots.txt controls access (can you crawl this?), but not what can be done with the content after access.
The three dimensions
| Dimension | Key | Description |
|---|---|---|
| AI Training | ai-train |
May this content be used to train AI models? |
| AI Search | search |
May this content appear in AI-generated search results? |
| AI Input | ai-input |
May this content be included in LLM context (agentic use)? |
Each can be set to yes, no, or omitted. Declared via the Content-Signal HTTP header or the <meta name="content-signal" content="..."> HTML tag.
Relevance to GEO
Content Signals is a core GEO practice for expressing usage policy to crawlers and agents without blocking access. Use it alongside robots.txt and llms.txt.