AI Principles & Terms

Our AI principles

• Be helpful to people everywhere, in their language.
• Be honest: say when unsure; label AI-generated media; never pretend to be human when asked.
• Be safe: refuse to help with serious harm, and test models before release.
• Protect privacy: minimise data, give people control over history.
• Keep people in charge: human review for important decisions about accounts, money and safety.
• Be fair and inclusive: evaluate quality across languages, countries and communities.

Acceptable use

Don’t use Xcojo AI to: create child sexual abuse material or sexualise minors; generate non-consensual intimate imagery or realistic sexual content of real people; harass, threaten or incite violence; plan or carry out crimes; create malware or attack systems; develop weapons capable of mass harm; impersonate real people deceptively; create disinformation intended to mislead about elections, health or emergencies; or infringe others’ rights. Don’t try to bypass safety systems.

Your output

Subject to these terms and the law, you may use what you create with Xcojo AI, including commercially. Similar output may be generated for others. You are responsible for checking output and for how you use it. Don’t present AI output as professional advice.

Your data

Prompts, files and responses are handled under the Privacy Policy. You can delete history, use Ghost chat, and opt out of quality improvement in JoinCTT settings.

Usage limits and credits

Plans include usage allowances per period. With usage credits enabled, extra usage is charged to your XCJ balance up to the limits you set; you can turn credits off any time.