html Terms of Service & Acceptable Use — Adverserial AI
ADVERSERIAL AI
Adverserial AI / Legal

Terms of Service
& Acceptable Use.

Powerful tools. Clear responsibilities.
The agreement for using our cybersecurity AI.

Version / 16 September 2026

Applies upon acceptance

Download terms PDF
01 / Authorization

Work within your permission.

Use CyberKimi and CyberGLM only for lawful activity, with the rights and authorizations your work requires.

02 / Human control

Review before you execute.

AI can be wrong or harmful. Validate outputs, constrain tools and keep people in control of consequential actions.

03 / Accountability

Responsibility follows conduct.

You are responsible for your decisions and misuse. Contractual protections remain subject to mandatory law.

A plain-language overview. The complete terms below control.

On this page — 22 sections