Ajay Pancholi of Carlsberg Britvic explains why AI must be treated as critical infrastructure, not innovation theatre, and how governance, systems thinking and safe experimentation enable accountable enterprise AI.
In this episode of Ctrl Alt Lead, Penny Horwood is joined by Ajay Pancholi, IT Infrastructure Lead at Carlsberg Britvic and founder of the AI Security, Ethics and Governance UK community, to explore why organisations need to rethink how they approach AI.
Pancholi argues that AI should be treated as critical infrastructure, not “innovation theatre”, warning that most enterprise risk sits outside the models themselves and lies within systems, data pipelines, identity and operations. Drawing on his infrastructure background, he explains why opaque vendor stacks, weak controls and poor integration create hidden risks, from data leaks to vendor lock‑in.
The conversation also tackles AI governance, reframing it as an enabler rather than bureaucracy, and looks at how organisations can design safe spaces for experimentation without inviting shadow AI. Looking ahead, Pancholi outlines what will force the shift from experimentation to accountable deployment, and what IT leaders should be doing now to prepare for greater scrutiny, regulation and scale.