Answers · AI output
Can I use Claude output commercially?
Last updated
TL;DR: Yes. Anthropic's Consumer Terms (Claude Free, Pro, Max and Ultra; effective October 8, 2025) assign you all of Anthropic's right, title and interest, if any, in Outputs, subject to your compliance with the terms. The Commercial Terms (API keys, Console, Team and Enterprise; effective June 17, 2025) go further: as between the parties the customer owns its Outputs, Anthropic may not train models on customer content, and Anthropic defends paid use against third-party IP claims, with exclusions. The limits are in what you do with it: no using Claude to build competing products or train AI models (outputs for training need Anthropic's prior authorization under the Usage Policy), no infringing others' rights, no passing AI-assisted work off without permission or attribution where that's required, and chatbots that talk to the public must say they're AI. Anthropic's assignment also can't create a copyright the law doesn't recognize.
What should I check, step by step?
1. Know which terms you're under. Claude Free, Pro, Max and Ultra (including Claude Code signed in with those plans) use the Consumer Terms; API keys, the Console, Team and Enterprise use the Commercial Terms. Claude Code's legal page lists the same split.
- Don't use evaluation or trial access commercially: the Consumer Terms limit evaluation use to personal, non-commercial use.
- Review Output before you ship it. Both sets of terms say Outputs can be wrong; the Commercial Terms make it your job to decide where human review is needed and to tell your users not to rely on factual claims without checking.
- Check Output for third-party material: quotes, code with license headers, logos, characters or a real person's likeness can still infringe, and infringing IP is a Usage Policy violation.
- Don't train models on Claude output without Anthropic's prior authorization, and don't use the service to build a competing product.
- Customer-facing bot or agent: tell users they're talking to AI at least at the start of each session or in the interface. High-risk advice or decisions (legal, medical, finance, credit, employment and others) need a qualified human reviewer and disclosure to the person affected.
- Need enforceable ownership (a logo, a book, code you'll license)? Add substantial human creative work and keep the drafts; in the US, material generated from prompts alone generally isn't protected by copyright.
- Keep a record: plan or contract, date, prompt and the exact Output you used.
Does a free Claude plan change the answer?
No. The Consumer Terms' Outputs clause doesn't depend on the plan: the same assignment covers Free, Pro, Max and Ultra. The difference is training: on consumer plans Anthropic may use Materials to train models unless you opt out in settings (and still may for feedback you give or content flagged for safety review); under the Commercial Terms it may not train on customer content.
- The consumer assignment is "if any": Anthropic hands over whatever rights it has, without promising the Output is original or protectable.
API, Team and Enterprise customers
Commercial Terms section B: the customer retains its Inputs and owns its Outputs, Anthropic disclaims rights it receives in customer content and assigns its rights (if any) in Outputs.
- Section K.1: Anthropic defends the customer against third-party claims that paid, authorized use of the Services or the Outputs infringes IP. Section K.3 excludes claims arising from your changes to Outputs, combining them with other technology or content, your Inputs, uses you knew or should have known infringe, practicing a patented invention in an Output, or trademark use of an Output in commerce. Read those exclusions before relying on it.
- Claude Code inside your own product (hosted sandboxes, agent infrastructure) needs the Commercial Terms, an unmodified binary, and each end user signing in with their own key or plan; you can't resell or route usage through consumer subscriptions.
Rules that apply whatever the plan
The Usage Policy shown on Oct 30, 2026 (effective Nov 12, 2026) applies to apps, Claude Code, API customers, cloud-provider customers and end users of products built on Claude.
- It prohibits infringing third-party IP, plagiarizing or submitting AI-assisted work without proper permission or attribution, using outputs to train an AI model without prior authorization, and bots that claim to be human.
- You may not use Anthropic's names or logos in a way that suggests affiliation or endorsement without written permission.
How do I ask Rightshold about this (free, no key)?
Add the MCP server https://www.rightshold.com/mcp to Claude, Cursor, ChatGPT, Grok or VS Code, then ask: “We used Claude through the API to write our product copy and part of our docs. Can we publish and sell them, and what do Anthropic's terms require?”
{
"tool": "ask_rightshold",
"arguments": {
"question": "Can I use Claude output commercially, and what do Anthropic's terms require?"
}
}- Free. Point source_url at the terms page you're under and Rightshold explains what it allows; it doesn't see your Anthropic account or chats.
- issue_rights_receipt (with an API key) binds a file's SHA-256 to the terms and intended use you record, signs it and commits it to the ledger, so you can later show what you relied on and when.
- check_rights_batch checks every text, image, font and package in a project in one call before you ship.
Rightshold does not grant or confirm Anthropic rights and doesn't analyze your content for copied material. Only Anthropic's terms, any contract you signed and the law decide. Not legal advice.
What are the sources?
- Anthropic Consumer Terms of Service (effective October 8, 2025)
- Anthropic Commercial Terms of Service (effective June 17, 2025)
- Anthropic Usage Policy
- Claude Code: Legal and compliance (which terms apply, offering Claude Code in products)
- U.S. Copyright Office: Copyright and Artificial Intelligence, Part 2: Copyrightability (January 2025)
Related
- Can I use ChatGPT output commercially?
- Who owns code written by an AI coding agent?
- Open weights vs open license: can I use model outputs commercially?
- How to check if AI-generated code copied open source code
Not legal advice. 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}}'