Answers · AI and artists
Does a "no AI use" notice on your art do anything legally?
Short answer: In the U.S., not much on its own. Copyright already protects your art, and a notice adds no new rights. People who merely view a post have not agreed to any contract, so a caption is unlikely to bind them, and whether training AI on publicly available work is fair use is being decided in court case by case. A notice still states your terms clearly and may help show what someone knew. The opt-outs with more practical and legal weight are machine-readable: robots.txt rules for AI crawlers, a text-and-data-mining reservation (which in the EU legally limits commercial mining of content made publicly available online), platform AI settings, and registering your copyright so you can enforce it in the U.S.
Checklist
- Keep the notice if you like: it is a clear statement of your terms, but don't rely on it alone.
- On a site you control, block AI training crawlers in robots.txt (GPTBot, ClaudeBot, Google-Extended, CCBot and others) and publish a machine-readable text-and-data-mining reservation such as W3C TDMRep.
- On platforms you post to, turn on any AI-training or data-sharing opt-out they offer, and read their terms: the license you grant the platform can matter more than your caption.
- Put your terms in your website's terms of use and in the metadata of the files you publish; for licensed work, put AI terms in the contract.
- Register important works with the U.S. Copyright Office: for U.S. works, registration is required before you can sue, and statutory damages and attorney's fees need registration before the infringement or within three months of publication.
- If you publish under a Creative Commons license, know it has no AI-specific clause: whether training is allowed turns on its ordinary conditions (such as NonCommercial) and on copyright exceptions. Add separate terms or a reservation if that matters to you.
- Keep dated originals and drafts so you can prove authorship and timing.
Why the U.S. and the EU differ
- U.S.: copyright protects your work, and there is no legal opt-out mechanism for AI training. Whether a particular training use is fair use is a fact-specific question that U.S. courts reached different conclusions on in 2025, and it is not settled.
- EU: the 2019 Copyright in the Digital Single Market Directive (Article 4) allows text and data mining of lawfully accessible works unless the rights holder has expressly reserved it, and for content made publicly available online the reservation should be made by machine-readable means. The EU AI Act requires general-purpose AI model providers to have a policy to respect those reservations.
- Other countries take other approaches; some allow mining for any purpose, some only for research.
License checks (computed)
Yes means the license grants the use, on the duties listed. Duties such as source_disclosure, share_alike and network_source_offer are the ones that conflict with keeping a product closed-source.
| Use | Answer | Why |
|---|---|---|
| Training a commercial AI model on CC BY 4.0 art (CC-BY-4.0) | Needs review | Creative Commons licenses grant rights to reproduce and adapt the material, with conditions that apply when it or an adaptation is shared. Whether training a model creates an adaptation, and how attribution applies to a trained model or its output, is not settled; Creative Commons notes that training may also fall under exceptions such as fair use. A human reviewer should confirm this use for your case. Duties: attribution, license_notice, state_changes, no_additional_restrictions. |
| Training a commercial AI model on CC BY-NC 4.0 art (CC-BY-NC-4.0) | No | CC BY-NC 4.0 prohibits commercial_use. Duties: attribution, license_notice, state_changes, no_additional_restrictions. |
Ask Rightshold (free, no key)
Add the MCP server https://www.rightshold.com/mcp to Claude, Cursor, ChatGPT, Grok or VS Code, then ask: “What does my portfolio site actually tell AI crawlers? Is AI training opted out?”
{
"tool": "read_ai_policy",
"arguments": {
"domain": "<your site's domain>"
}
}- It reads robots.txt for 14 AI crawlers and user-fetch agents, ai.txt and W3C TDMRep, and summarizes whether AI training is opted out. These are signals, not a license.
- check_license tells you what a Creative Commons or other license you publish under allows for model_training.
Rightshold reads the opt-out signals and licenses a site publishes. It does not enforce them, and it does not decide whether a given AI use is fair use.
Sources
- U.S. Copyright Office: Copyright and Artificial Intelligence
- 17 U.S.C. § 412: registration as a prerequisite for statutory damages and fees
- Directive (EU) 2019/790 (Copyright in the Digital Single Market), Article 4
- Regulation (EU) 2024/1689 (AI Act), Article 53
- W3C TDM Reservation Protocol (TDMRep)
Related
- Can I use AI-generated images commercially?
- Can I use CC BY-NC 4.0 commercially?
- Who can file a DMCA takedown?
Not legal advice. Updated 2026-10-28. JSON · Markdown · All answers
Get a free API key in one call
No card and no email needed. 100 free credits; the license and policy checks stay free without a key.
curl -s -X POST https://www.rightshold.com/v1/accounts/signup -H "Content-Type: application/json" -d '{"accept_terms":true}'
export RIGHTSHOLD_API_KEY=rs_... # the api_key from the reply (shown once)
curl -s https://www.rightshold.com/v1/licenses/check -H "Authorization: Bearer $RIGHTSHOLD_API_KEY" \
-H "Content-Type: application/json" -d '{"license":"MIT","intended_use":{"commercial_use":true}}'