Acceptable Use Policy

This Policy sets out how the Service may and may not be used. Words defined in the Terms of Service have the same meaning here, and the “cyankiwi agent” is the workspace agent.

Section 1 lists what is forbidden on every part of the Service. Sections 2 to 4 set out what each service may be used for.

1. Forbidden on every service

You must not use the shared inference API, a Deployment or the cyankiwi agent, or allow them to be used, to:

Break the law. Do anything unlawful, or infringe anyone’s rights.

Sexual content involving children. Create, request or distribute child sexual abuse material, or sexualise a minor in any way, including through wholly synthetic images. You must also not develop, train, adapt or supply any model, program or service designed or optimised to produce such material, or use the Service to generate training data for one.

Sexual content without consent. Create, request or enable intimate or sexual images of a person without their consent, including synthetic and “deepfake” images that appear to depict an identifiable person, or build or supply any tool for that purpose.

Sexually explicit content. Generate pornographic or sexually explicit material.

Weapons capable of mass casualties. Develop them, or seek or provide meaningful help towards chemical, biological, radiological or nuclear harm.

Terrorism and violent extremism. Promote, encourage or facilitate terrorism, create or distribute terrorist material, or generate information likely to be useful to a person preparing an act of terrorism.

Attacks on systems. Create or deploy malware, or attack, disrupt or access computer systems or critical infrastructure without authorisation. Authorised testing of your own systems, or of systems whose owner has instructed you, is permitted.

Fraud, harassment and impersonation. Defraud, defame, harass, threaten or impersonate anyone.

Deception about authorship. Present AI-generated content as written by a person where a reader’s decision would reasonably depend on believing it was, such as in journalism, reviews, political communication, academic work or legal proceedings.

Prohibited AI practices. Use the Service for any practice prohibited by Article 5 of the EU AI Act, including manipulation or exploitation of vulnerable people that causes significant harm, social scoring, emotion recognition at work or in education, untargeted scraping of facial images, and biometric categorisation to infer protected characteristics.

Circumvention. Circumvent rate limits, access controls or safety measures, or try to extract model weights or parameters through systematic querying.

2. The shared inference API

You may use the shared inference API for any purpose that section 1 does not forbid. If your product lets people interact with AI-generated output, you must make sure they know they are interacting with an AI system, unless that is obvious. If you publish AI-generated images, audio or video that resemble real people or events, or AI-generated text informing the public on matters of public interest without human editorial responsibility, you must disclose that it is AI-generated. Do not remove any marking or provenance information present in AI-generated output.

3. Deployments

You may use a Deployment for any purpose not forbidden by section 1, on the same conditions as section 2. You must not deploy a model that is designed or adapted for a purpose section 1 forbids, and you are responsible for having the right to use the models you deploy.

4. The cyankiwi agent

The cyankiwi agent may be used only in relation to cyankiwi services: choosing, planning, starting, stopping and troubleshooting Deployments, and questions about the Service, its API, API keys and Credits. You must not:

  • use it as a general-purpose assistant for anything unrelated to cyankiwi services;
  • ask it to deploy a model that section 3 does not allow;
  • try to make it act outside its tools or bypass its safeguards, including by prompt injection or jailbreak attempts.

Requests to the agent that break this Policy are refused.

5. Reporting misuse

Report misuse of the Service to abuse@cyan.kiwi.