Organizations embrace AI but lack proper governance over development
According to new research 93 percent of firms in the UK today use AI in some capacity, but most lack the frameworks to manage its risks and don’t integrate AI governance into their software development processes.
The study from Trustmarque shows only seven percent have fully embedded governance frameworks to manage AI risks. In addition a mere four percent consider their technology infrastructure fully AI-ready, and just eight percent have integrated AI governance into their software development lifecycle.
Organizations still largely rely on legacy software development processes which haven’t been updated to account for AI-specific risks such as model bias or explainability gaps. Only 28 percent apply bias detection during testing, and fewer still (22 percent) test for model interpretability.
Infrastructure and tooling also present barriers to governance according to the IT leaders surveyed. Just four percent of organisations say their data and infrastructure environments are fully prepared to support AI at scale. Registries, audit trails, and version control for AI models are often manual or missing entirely.
Responsibility for AI oversight is fragmented too. Of those surveyed nine percent report alignment between IT leadership and governance, while 19 percent say there is no clear owner for governance activity. Executive engagement remains low too, with most AI governance driven at the departmental level rather than through strategic leadership.
Seb Burrell, head of AI at Trustmarque says:
Our report shows that AI adoption is outpacing governance, It’s a clear gap -- 93 percent of organisations are using AI but only seven percent have fully embedded governance frameworks.
Right now, systems and processes haven’t kept up with the speed of innovation. Development teams lack proper tooling and infrastructure, and the issues are compounded by the lack of management buy-in for building robust governance systems.
Governance is also often seen as a constraint, but our findings suggest otherwise. Organisations that have adopted AI governance are seeing real, tangible benefits -- faster deployments, stronger accountability, and reduced manual review cycles. It’s a vital support function to enable responsible, scalable AI.
You can get the full Trustmarque AI Governance Index 2025 from the company’s site.
Image credit: Michael Borgers/Dreamstime.com