Back to home
Services
Artificial intelligence audit
We review whether the artificial intelligence systems you already run meet the European regulation and GDPR. We do it from engineering, not from a template.
What we review
- Risk classification
- Which category each system falls into under the European regulation, which determines everything else.
- Lawful basis and data
- What data goes in, under what basis, and what happens to the rights of the people affected.
- Technical documentation
- Whether what you have written would hold up if somebody asked for it, and what is missing.
- Human oversight
- Whether a person genuinely decides, or merely approves what the system already decided.
Why we do this
We build an artificial intelligence product that has to stand up in a regulated sector. We have had to answer these questions about our own system before answering them about anyone else's. The output is not a generic report: it is a list of what to change, in what order and at what effort.
What systems do you have running?
With a rough inventory we can scope the audit and say where we would start.