New Prime Minister Andy Burnham’s first cabinet has produced a genuine first for British government. Kanishka Narayan, previously the junior minister for AI and online safety under Liz Kendall, has been promoted to a dedicated Minister for AI role that will attend cabinet, the first time this brief has carried that weight. He is not new to the job. He has spent nearly a year in the junior version of it, so whatever comes next builds on a record rather than starting from scratch.
What makes this more than a straightforward promotion is what happened to the department around him. The Deparment for Science, Innovation and Technology has been scrapped. Its remit has been split three ways, with the bulk merging into a new Department for Business, Innovation, Science and Trade under Jonathan Reynolds, while the Culture Department absorbs digital transformation and online harms. AI gets its own political voice for the first time in the UK, at the exact moment the department that used to house it has been broken into pieces and distributed elsewhere. Boards should read that as a mixed signal rather than a clean upgrade. A dedicated minister raises AI’s profile. A dismantled department raises the question of who is actually joining the pieces back together.
That question was already live before this reshuffle. Lord Chris Holmes has spent three years arguing for a single AI authority with real enforcement powers, pressing the case again in the Lords as recently as June. Government has consistently preferred to work through existing sector regulators rather than create anything new. Narayan now inherits that unresolved argument directly, and the DSIT breakup makes the case for coordination harder to dismiss, not easier.
Sitting right next to it is a second question this government has spent over a year avoiding. Ministers have run consultation after consultation on AI and copyright, dropped their own preferred model, and called the result a genuine reset moment without ever coming up with an answer. Creators are losing income while the uncertainty continues, and any business building on training data or licensing arrangements needs a decision, not another round of engagement.
Underneath both of these sits the problem your organisation actually lives with day to day. Skills funding and innovation announcements are not the same as people using AI well in their jobs. The application gap between having access to the tools and using them effectively does not close through a ministerial appointment, however senior. Narayan’s brief will need to reach into adoption and practice or it will only ever address half the problem.
None of this is helped by a landscape that already includes the AI Security Institute, the Regulatory Innovation Office and a dozen sector regulators, now joined by a business department absorbing most of DSIT and a culture department taking the rest. A cabinet level minister sitting across all of that needs to be a coordinating force, not one more voice in an already crowded room.
And whatever gets built needs to hold onto a distinction this government keeps blurring. Assurance tells you a system passed a test on a given day. Governance is the ongoing judgement about whether an organisation is using AI responsibly, and that judgement was never going to come from a certificate.
Narayan’s own comment on his appointment, describing AI as likely the most significant technology in human history, tells you where his instincts sit. Whether that translates into the coordination and clarity boards actually need is the thing worth watching over the next few months, not the announcement itself.
