Arkeo AI Acceptable Use Policy

Welcome to Arkeo AI. Your agents can do a lot on your business's behalf, so every customer agrees to use them responsibly. This Acceptable Use Policy sets out what you and your agents may not do on Arkeo, and what Arkeo does when these rules are broken.

Effective date: October 23, 2026
Version: 2026-10-23

This Acceptable Use Policy ("AUP") is part of the Arkeo AI Business Terms. It applies to you, your Authorized Users, and every Agent and Build running on your Machine. Capitalized words have the meanings given in the Business Terms.

You are responsible for making sure your Builds and Instructions do not cause an Agent to break this AUP. This holds even if an Agent created or changed the Build itself. "Use" below includes anything you, an Authorized User, or an Agent does through the Services.

These rules do not replace the law, your professional duties, or the terms of your Connected Systems. Where those set a higher standard, follow them.

1. Protect people

You may not use the Services for:

• threats, intimidation, harassment, stalking, bullying or defamation
• promoting or facilitating suicide, self-harm or eating disorders
• sexual violence, or non-consensual intimate content
• terrorism, violent extremism, or violence, including hate-based violence
• developing, acquiring or using weapons, including chemical, biological, radiological, nuclear or explosive weapons
• illegal goods, services or activities
• real-money gambling
• selling or providing age-restricted goods, services or content, such as alcohol, tobacco, firearms, gambling or adult content, to minors
• content that sexualizes, grooms, exploits or endangers anyone under 18, under any circumstances. Arkeo reports apparent child sexual abuse material to the National Center for Missing and Exploited Children.

2. Respect privacy

You may not use the Services to:

• collect, aggregate, monitor, profile or distribute anyone's private or sensitive information without authorization and a lawful basis
• build facial recognition databases without consent, or carry out real-time biometric identification in public spaces
• use anyone's name, likeness, image or voice without consent in a way that could mislead people about who they are dealing with
• infer the emotions of employees or students, except where necessary for medical or safety reasons
• score, rank or classify people based on social behavior, personal traits, biometric data, or inferred sensitive attributes such as health, race, religion, sexual orientation or political views
• predict whether a person will commit a crime based on profiling or personal traits

3. No deception, fraud or abuse of outreach

You may not use the Services for:

• fraud, scams, phishing, or deceptive practices
• impersonating a person or organization, or misrepresenting who an Agent is acting for
• spam, or calls, texts or emails that violate outreach laws, ignore opt-outs, or lack required consent
• fake reviews, fake testimonials, or fake engagement
• presenting AI-generated content as written by a person where that would mislead
• hiding that an Agent is AI where the law requires disclosure
• chaining or sequencing automated actions across systems in a way designed to hide their automated nature from the people affected
• price-fixing, bid-rigging, market allocation, coordinating prices or output with competitors, or other anticompetitive conduct
• political campaigning, lobbying, election interference or voter suppression
• academic dishonesty

4. High-Impact Actions need a person

You may not build or allow Agents to make High-Impact Actions without a person you trust reviewing and approving each one before it is final. High-Impact Actions are defined in Section 5 of the Business Terms. They cover decisions about employment, pay, credit, insurance, housing, education, legal rights, health and safety, and actions that move money, sign contracts or file with government agencies.

You may not present an Agent as a licensed professional, or use Agents to give advice that requires a license without a licensed professional involved.

5. Protect systems

You may not use the Services to:

• access, scan, test, attack or disrupt any system, network or account you are not authorized to access
• create, store or distribute malware, or carry out denial-of-service or other cyberattacks
• infringe or misappropriate anyone's intellectual property
• bypass, disable or work around security controls, rate limits or safeguards on the Services or on any Connected System
• access Arkeo's systems, other customers' machines, or parts of the Machine Arkeo has not made available to you
• run security testing against the Services without Arkeo's written permission
• violate or get around another service's terms, CAPTCHAs, paywalls or access controls, or use anyone's credentials without authorization
• control physical systems or equipment in a way that could endanger people or break the law
• mine cryptocurrency, or use the Machine for workloads unrelated to running your Agents

6. Protect the platform and model providers

You may not:

• resell the Services, or use them to give anyone direct access to AI models
• use the Services or Output to build, train or improve an AI model or product that competes with Arkeo or its model providers, including by creating training data, labels, scores or rankings
• attempt to extract model reasoning, system prompts or model weights through circumvention
• copy or reverse engineer the Arkeo platform

7. Your own responsibilities

You must:

• give access only to Authorized Users you trust, and remove access when it is no longer appropriate
• give each Agent only the access it needs
• check what your Agents build and change, and pause any Agent that acts in ways you did not intend
• keep your own backups of your Connected Systems
• obtain the rights, notices and consents needed for the data you connect and the outreach your Agents carry out
• comply with US export control and sanctions laws

8. Enforcement

Arkeo may investigate suspected violations. Arkeo may pause or remove Agents or Builds, disable specific Agent capabilities, refuse or block Actions, restrict access, suspend or terminate your account, and report unlawful activity to authorities. Arkeo is not obligated to monitor your Builds, Instructions or Actions. When Arkeo does act, it does not take on responsibility for them.

Report misuse or security issues to support@arkeoai.com. If you believe Arkeo enforced this AUP in error, contact the same address.

9. Changes

Arkeo may update this AUP as described in Section 19 of the Business Terms.