Responsible AI statement
How I use AI in my own work and how I help organisations use it. This is a working statement, informed by responsible AI guidance including the NIST AI Risk Management Framework. It is not a certification.
Human-directed by design
AI drafts. People decide. Every workflow I design places a person at the points where a wrong output would cost something, and records that approval so the result is traceable.
Consequential decisions about people, money, safety or reputation remain with accountable humans. I do not build autonomy a system has not earned through demonstrated reliability on the specific task.
Learning first
I teach people to specify bounded work, recognise uncertainty, verify outputs and stop or escalate. Independence is earned task by task. Programmes include assessed independent work and a follow-up, because enjoyment is not evidence of adoption.
Data and boundaries
Every engagement begins by identifying sensitive information, approved tools, access boundaries, human review points and retention requirements. I do not recommend automating a workflow until its risks are understood.
Creative work
Where I use AI in production I direct it, review it and say what was made how. Characters, continuity and the final judgment stay human-directed. Generated placeholder media on this website is labelled as a concept preview and is never presented as delivered client work.
Children and learners
Learners remain the authors of their work. Activities for schools are designed so that no child's personal information needs to leave the school, and educators review every step.
