TERMS
Engagement terms.
Short, on purpose. The engagement letter always wins.
01
What we sell
Adversarial evaluation of AI systems and related documentation, as described on this site and scoped in a written engagement letter: Agent Red Team, EU AI Act Art. 15 Evidence Pack, Continuous Assurance.
02
How it starts
A signed engagement letter defines scope, authorized targets, contacts and schedule. We begin only after written authorization. Verbal scope is not scope.
03
Prices and payment
Prices as published; engagement letters may deviate. Invoices are payable within 14 days. Continuous Assurance renews monthly, cancellable to the end of any month.
04
What we guarantee
Method diligence, not outcomes. We do not guarantee the absence of vulnerabilities or compliance of your system. Findings reflect the state of the tested system during the engagement window.
05
Confidentiality
Your data and our findings are confidential by default. We publish nothing about an engagement without your written permission, and nothing at all that could identify non-public targets.
06
Your responsibilities
You warrant that you own or are authorized to test every system in scope, and that designated test accounts are yours.
07
Liability
Liability is limited to the fees of the affected engagement, except for intent, gross negligence, personal injury and statutory product liability.
08
Consumer withdrawal
If you act as a consumer (which includes sole traders under German law), statutory rights — including the right of withdrawal for distance service contracts — apply in addition to these terms; nothing here contracts them away.
09
Law and courts
Spanish law, excluding the CISG. Courts of Palma de Mallorca, unless mandatory consumer law gives you a different forum.
Version 2026-11-06 · Changes apply to future engagements only.