What we use AI for
- Writing. A large language model assistant helps draft posts, Lab Notes, course material and this site. George Vossos reviews everything published and is accountable for it.
- Code. A large language model assistant helps write this site and the Evaluator Lab code. The Lab code is public at github.com/gvossos/evaluator-lab, so anyone can check it.
- Labs. Lab agents are teaching examples for fictional organisations, run on test data, never client data.
- Client work. The AI tools used on a client's work are named in the engagement letter or statement of work.
Who checks
- At our Evaluator Labs, George builds the agent, the room judges it, and the builder never signs the sign-off record.
- Statements about the law on this site are checked against the regulator's own source, dated "as at", and reviewed by an external legal adviser before publication.
Decisions about people
We have not arranged for any computer program to make, or do something substantially and directly related to making, decisions about you that could reasonably be expected to significantly affect your rights or interests. Enquiries and sign-ups are read and followed up by a person. See also our privacy policy.
Independence
We don't review agents that Netlifestyle AI or a related party built.
Changes
8 Oct 2026 first published.
General information, not legal advice. For what the rules mean for the agents your firm runs, read AI transparency for firms using agents: Australia and the EU.