Universal standards
Violate law or engage in or promote illegal activity.
Compromise computer or network systems — including unauthorized access, scanning, or exploitation of systems you do not own or lack written authority to test; developing or distributing malware, ransomware, or credential-harvesting tools; or evading security controls.
Compromise critical infrastructure or develop weapons.
Violate privacy or identity rights — including surveilling, tracking, or profiling individuals without a lawful basis, or processing personal data through the Services without the rights to do so.
Infringe intellectual property or misappropriate trade secrets.
Deceive — impersonate any person or organization, misrepresent AI-generated content as human-authored where disclosure is required, or generate content designed to defraud.
Harm children, generate sexual content involving minors, or generate non-consensual intimate imagery.
Harass, threaten, or incite violence or hatred.
Abuse the platform — circumvent rate limits, resell access, share credentials, or use the Services to train a competing model.
Use prohibited AI practices — including deploying subliminal, manipulative, or deceptive techniques that cause significant harm; exploiting vulnerabilities due to age, disability, or social or economic circumstance; social scoring; predicting criminal behavior based solely on profiling; untargeted scraping of facial images; inferring emotions in workplaces or schools; or biometric categorization to infer sensitive characteristics — where the law prohibits them.
1
Ensure a qualified human professional reviews the output before it is relied upon or disseminated;
2
Disclose to affected individuals that AI was used to help produce it;
3
4
Enforcement
