Acceptable Use Policy

This Acceptable Use Policy applies to everyone who uses Chernion, across the API, the hosted chat, and the coding command line agent. It is part of the Terms of Service. If you break this policy, we may act under the enforcement section below and under the Terms of Service.

Every model you reach is run by an upstream provider, and that provider's usage policy applies to your use of its model. You must read and follow them. Where a provider policy is stricter than this one, the stricter rule controls. At the date of this draft the relevant policies are:

• The OpenAI usage policies, for the GPT and Codex families.
• The Anthropic usage policy and acceptable use policy, for the Claude, Opus, and Fable families.
• The Google generative AI prohibited use policy, for the Gemini family.

A use that any of these providers prohibits is also prohibited on Chernion, even if it is not spelled out below.

You must not use the service to create, request, store, or distribute, or to attempt any of the following:

• Anything illegal, or anything that promotes or facilitates an illegal act.
• Child sexual abuse material, or any sexual content involving minors, and any attempt to sexualise a minor.
• Content that exploits, endangers, or sexualises a real person without consent, including intimate images shared without consent.
• Harassment, bullying, threats, stalking, or content that incites or promotes violence, self harm, or terrorism.
• Hateful content that demeans or attacks people on the basis of a protected characteristic.
• Malware, ransomware, exploits, or other code, instructions, or tooling whose purpose is to break into, damage, or disrupt a system, network, or account, and instructions for building weapons that can cause mass harm.
• Fraud, phishing, scams, spam, impersonation, fake reviews, or other deceptive or manipulative practices, including coordinated inauthentic behaviour.
• Disinformation presented as fact, or content meant to deceive people about who created it where that deception causes harm.
• Infringement of intellectual property, trade secrets, privacy, or other rights, including sending or extracting personal data without a lawful basis.
• Regulated professional advice, such as legal, medical, or financial advice, presented as coming from a qualified professional when it does not, or used to make high stakes decisions about a person without qualified human review.

On top of the content rules, you must not abuse the gateway itself. You must not:

• Try to get around rate limits, balance checks, quotas, safety systems, or content filters, or probe for ways to do so.
• Resell, redistribute, or sublicense access to the models in a way that breaks an upstream provider's terms, or use the service to build or operate a competing reseller in breach of those terms.
• Extract model weights, training data, or system prompts, or use the service to reverse engineer or copy a model, except to the narrow extent the law allows and no contract forbids.
• Share, sell, or publish your API keys, or let an unauthorised person use your account or balance.
• Use bots, scripts, or other automation to abuse free or promotional features, to create accounts in bulk, or to place load that harms the service or other customers.
• Run security testing, scanning, or penetration testing against the service or the upstream providers without our prior written permission.
• Interfere with, overload, or disrupt the service or the networks and systems behind it, or bypass any access control.

You must not use the service as the sole basis for a decision that has a legal or similarly significant effect on a person, or for a use where an error could cause death, personal injury, or serious harm, unless a qualified person meaningfully reviews the output and appropriate safeguards are in place. You are responsible for the safeguards around your own use.

If you break this policy, we may remove or restrict content, throttle or suspend your access, revoke API keys, close your account, and, where you have seriously or repeatedly breached this policy, forfeit the remaining prepaid balance to the extent the law allows. We choose the response that fits the breach, and urgent risks may be acted on at once.

We may report unlawful content or conduct to the upstream providers and to the authorities, and we cooperate with lawful requests from competent authorities. We may preserve and disclose information where the law requires it or allows it to protect people or the service.

If you see use of Chernion that breaks this policy, tell us at [email protected] with enough detail for us to look into it.