1. Intent
This policy describes what tAI, tCode, and the Artfical API are built to support, and the specific uses that fall outside that regardless of who is asking. It's written to work alongside, not replace, the judgment described throughout the security docs.
2. Universal prohibitions
- Developing or deploying weapons capable of causing mass casualties, including chemical, biological, radiological, or nuclear weapons.
- Generating child sexual abuse material, in any form.
- Unauthorized access to a computer system, or developing malware intended for unauthorized use.
- Coordinated inauthentic activity intended to manipulate an election or deceive the public at scale.
- Any use that violates applicable law in the jurisdiction where it takes place.
3. High-risk uses require human oversight
Uses involving legal, medical, or financial decisions that materially affect a person, or automated decisions about employment, credit, housing, or similarly consequential outcomes, require a qualified human to review the output before it's acted on. tAI is not authorized to make these decisions autonomously.
4. Security research
Legitimate security research, penetration testing you're authorized to perform, and CTF-style educational work are permitted uses, evaluated under the standard described on the Cybersecurity misuse page rather than refused outright.
5. Automated and high-volume abuse
Automated use that circumvents the rate limits described in the applicable terms, or that's designed to degrade the Services for other users, is prohibited regardless of whether the underlying requests would individually be permitted.
6. Enforcement
A violation of this policy is handled under the same enforcement framework described in the security docs: a warning through to account suspension depending on severity, with the most severe categories enforced without a warning step.
7. Reporting a violation
If you believe someone is using tAI in violation of this policy, report it through the contact details at ai.artfical.com/about.
8. Changes to this policy
Material changes are announced at tai.artfical.com/announcements. Prior versions are preserved at version history.
See version history for prior versions of this document, and the other pages under Terms and policies in the footer below for related terms.