Acceptable Use Policy

Everything your agents send to your AI provider must also follow that provider's own usage policy. For Anthropic, that is the Anthropic Usage Policy, which among other things requires agents that talk to the public to say they're AI, and a qualified person to review AI recommendations in high-risk areas before they're acted on. Breaking your provider's rules can get your provider account suspended.

Don't use the Service, or let your agents use it, to:

• Break the law or help someone else break it.
• Create or share child sexual abuse material, or sexualize minors in any way. We report it to the authorities.
• Threaten, harass, stalk or incite violence against anyone, or promote terrorism or violent extremism.
• Discriminate against people based on protected traits.
• Run scams, phishing, fake reviews or testimonials, or impersonate a person or business.
• Spread disinformation, run deceptive influence campaigns, or interfere with elections.
• Infringe someone's copyright, trademark, privacy or publicity rights.
• Help make weapons, including biological, chemical, nuclear or radiological weapons.
• Watch or track people without their permission or a legal right to.

If your agents email, message, post, call or fill in forms:

• Follow anti-spam and telemarketing laws (for example CAN-SPAM, TCPA, CASL and EU rules). No unsolicited bulk messages, no bought or scraped lists where the law forbids them, a real sender, and a working unsubscribe in marketing messages.
• Tell people they're dealing with AI when they ask, and wherever the law or the platform requires it. Label AI-made content where required.
• Follow each platform's rules on automation and bots.
• Disclose paid or material relationships in endorsements, as the FTC's rules require.

Don't let agents make decisions on their own that have legal or similarly serious effects on people (for example about credit, jobs, housing, insurance, education, health care or legal status). A qualified person must review those decisions before they're acted on, and the people affected must be told AI was involved where the law or your AI provider requires it. Don't present AI output as legal, medical, financial or tax advice.

Don't use the Service, or let your agents use it, to:

• Try to reach another customer's AI company or data, or any part of the Service you aren't allowed to use.
• Try to get around your AI company's isolation, resource limits or security, or reach our servers, internal networks or cloud metadata.
• Spread malware or other harmful code.
• Attack or disrupt any system: denial of service, scanning or testing systems without the owner's permission, password guessing or credential stuffing.
• Run proxies or relays for others, or hide where traffic comes from to dodge blocks or the law.
• Use logins, keys or accounts that aren't yours or that you aren't allowed to use.

Found a security problem in OpsHarbor? Email [email protected] with the details. Test only as much as needed to show the problem, don't touch other customers' data, and give us a reasonable time to fix it before you share it. We won't take action against good-faith reports that follow these rules.

• Stay within your plan's limits. Don't get around them with extra accounts or fake sign-ups.
• Use your AI company to run your AI company. No cryptocurrency mining, general file hosting, or selling our computing power to others.
• Don't slow the Service down for other customers. We may limit heavy workloads to protect everyone.

Respect each website's terms, robots rules and rate limits. Don't collect data from behind logins or paywalls you aren't allowed past, and don't collect personal data without a lawful reason.

• Don't store or create content you don't have the right to use.
• Don't use OpsHarbor or AI output to build or train a competing AI model or service.
• Don't resell the Service as a hosting product without our written permission. Agencies may run AI companies for their own clients under the Agency plan.
• Don't suggest OpsHarbor endorses you or your agents' work.

Depending on how serious it is, we may ask you to fix it, switch off the agent, content or connection involved, pause agents or limit features, suspend your AI company (right away if needed to stop ongoing harm, such as spam or an attack), close your account for serious or repeated problems, and report illegal activity to the authorities. We'll tell you why as soon as we can. To check a report, we may look at the relevant part of your AI company, only as much as the check needs.

To report spam, attacks or harmful content coming from an OpsHarbor customer, email [email protected] with as much detail as you can (links, message headers, times, IP addresses). For copyright complaints, see section 15 of the Terms of Service.

We'll update this policy as the Service and the law change, post the new version here, change the date at the top, and email customers about important changes.