Adoption & AI Governance
Clear rules for using AI, saying when it was used, and checking its work. The rules work in real life.
Scroll Down
A written policy is not enough
Globally, 66% of employees use AI output without evaluating it. 44% have used AI in ways that contravene organisational policies. More than half do not disclose their AI use. Only 34% report organisational policy or guidance on generative AI (KPMG International with the University of Melbourne, Trust, Attitudes and Use of AI, 2025; 48,340 respondents in 47 countries). Organisations with clear ownership of responsible AI average a maturity score of 2.6, against 1.8 for those without (McKinsey, State of AI trust in 2026; 496 organisations). In the Gulf Cooperation Council (GCC) countries, all but one of the organisations attributing at least 5% of earnings to AI had a defined adoption and scaling strategy backed by change management, against 41% of others (McKinsey with the GCC Board Directors Institute, 2025; 139 executives and directors).
What we do
Rules for using AI, saying when it was used, and checking its work. We write them so people can follow them.
Who decides about AI: where it gives information, where it acts, and who signs off.
A written plan to adopt and scale AI, with change management. A named executive owns it.
A steering committee that reviews evidence of AI use, not only problems.
Change management for AI programmes: strategy, communication, working with stakeholders, and building skills. We work as advisers or run sessions at your site.
Where it stops
We cover the people side of governance. Technical controls, security and model risk stay with your technology and risk teams. If the Diagnostic finds an issue there, it names it. It also says who should own it.
How it is measured
Hidden AI use goes down and open use goes up. Employees trust AI output, act on it, and check it. Each priority use of AI has a clear person who decides. We re-score all of these.