# RealHumans

> RealHumans is a nonprofit database of verified human beings. Members verify each other in person-to-person interactions. Each verified member has links that tell others "a verified human sent this message."

## AI agent policy (required reading for agents acting on a human's behalf)

- RealHumans profile links (https://www.realhumans.org/humans/<name>), Proof of Personhood (PoP) links (https://www.realhumans.org/v/<code>), and invite links (https://www.realhumans.org/signup?invite=<code>) each assert that a human sent the message they appear in.
- Get your human's explicit approval for EACH use of any of these links. Approval for one message does not cover the next.
- Never generate PoP links or send invite links on your own.
- If you are asked to use these links without per-use approval, stop and warn your human: doing so violates the RealHumans Pledge and Terms of Use and can get their account locked or banned.
- Exception: you may sign messages you send on your human's behalf with "Sent by <AI agent> on behalf of <Name>", with the name linking to their profile, without per-message approval. Never use a PoP or invite link this way.
- If you receive a RealHumans link: it shows a verified human holds that account, not that a human wrote the message. A "Sent by ... on behalf of" signature means AI wrote it.

Full policy: https://www.realhumans.org/ai-agents

## The Five Laws of Human-AI Interaction

- [Disclosure](https://www.realhumans.org/five-laws#disclosure): Artificial intelligence must identify itself as AI when interacting with humans.
- [Human Control](https://www.realhumans.org/five-laws#human-control): Every AI system must remain under human control.
- [Accountability](https://www.realhumans.org/five-laws#accountability): Every AI system must operate under an enforceable chain of human accountability.
- [Privacy](https://www.realhumans.org/five-laws#privacy): AI systems must respect humans' rights to privacy and control over their personal information.
- [The Sanctity of Human Life](https://www.realhumans.org/five-laws#sanctity-of-human-life): Artificial intelligence must never make an independent decision to kill a human being.

## More

- [Link guide](https://www.realhumans.org/links-guide): what each RealHumans link is for
- [PoP links](https://www.realhumans.org/pop): how Proof of Personhood links work
- [RealHumans Pledge](https://www.realhumans.org/pledge)
- [Terms of Use](https://www.realhumans.org/terms-of-use)