Acceptable use policy

Haarness Cloud gives you real machines, network access and AI models. This policy sets the limits that keep them safe for everyone. It is part of the Terms of service and applies to everything you run, store or publish on the Service, including what your agents do on your behalf.

You may not

Break the law or harm people

• Store, generate or distribute unlawful content, including child sexual abuse material, content that infringes intellectual-property rights, or personal data processed without a lawful basis.
• Harass, threaten, defraud or impersonate people, or generate content intended to deceive people about who produced it where that causes harm.
• Use the Service in breach of export-control or sanctions laws, or for anyone subject to them.

Attack systems

• Access, probe or scan systems, networks or accounts you are not authorised to test — including other tenants' workspaces, and our control plane, gateway, router or infrastructure beyond the interfaces we publish. Good-faith security research is welcome under our Security policy.
• Develop, host or distribute malware, run command-and-control infrastructure, phishing pages or credential harvesting.
• Launch denial-of-service attacks, or send spam or bulk unsolicited messages.
• Circumvent the Service's isolation, metering, rate limits, billing or usage restrictions.

Abuse the resources

• Mine cryptocurrency, or run workloads whose main purpose is to consume compute rather than develop software.
• Run open proxies, VPN exit nodes, or anonymising relays for third parties.
• Resell raw model access or share router keys with others; router and API keys are for your own applications and your organisation.
• Create multiple accounts to obtain extra free allowances or evade a suspension.

Misuse AI models

• Use models in breach of their provider's usage policies, which also apply to you when you use those models through the Service.
• Use the Service to build weapons capable of mass casualties, or to generate content that sexualises minors.
• Use model output to make decisions with legal or similarly significant effects on individuals without human review.

Automated workloads

Agents, schedules and API clients must respect the rate limits on your keys and plan. You are responsible for what an agent does in your workspace, including commands it runs with the permissions you gave it.

Marketplace

Resources you publish must work as described, must not collect data or run code their description does not disclose, and must respect the licenses of the code they contain. We may remove a listing that breaks these rules.

Enforcement

If we become aware of a breach we may, depending on its seriousness: ask you to fix it; remove the content; stop a workspace or revoke a key; suspend the account; or terminate it. Where it is safe and lawful to do so, we tell you what we did and why, and give you a chance to respond. We may report illegal content to the authorities.

We do not routinely inspect workspace contents. We act on reports, on signals from our own abuse and billing systems, and when the law requires it.

Reporting abuse

Report a breach of this policy — including illegal content — to [email protected]. Tell us where the content or activity is and why it is a problem; for suspected illegal content, explain why you believe it is illegal and give your name and email unless the report concerns child sexual abuse material. We acknowledge reports, act on them promptly, and tell you what we decided.

If we remove your content or restrict your account because of a report, you can contest the decision by replying to our notice or writing to [email protected].