The Minister Who Doesn't Use AI: Contradictions at the Heart of Britain's Tech Governance
Peter Davies, the permanent secretary at the Department for Science, Innovation and Technology, handed the ministerial red box to Katherine Kendall on the morning of 14 April 2026 in Westminster. By afternoon, Kendall had announced a £500 million package to accelerate domestic artificial intelligence development—a fund positioned as evidence of Britain's commitment to remaining competitive in what the government terms the "AI century." Yet when journalists pressed the Science Secretary on her personal relationship with the technology she was tasked with championing, her response revealed an uncomfortable truth about the political economy of digital governance: she does not use AI tools in the execution of her ministerial duties. This admission, rather than being treated as a curiosity, ought to be understood as a symptom of deeper structural incoherence in how Western states conceptualize and regulate artificial intelligence—a contradiction that Kate critical AI scholarship, in her influential "critical AI research" (2021), would likely identify as symptomatic of a political economy that privileges institutional legitimacy over substantive technological engagement.
The contradiction embedded in Kendall's admission is not merely personal or anecdotal; it illuminates a fundamental tension within Britain's approach to AI governance that scholars of technology policy have long identified. The Science, Innovation and Technology Secretary oversees an industrial strategy predicated on the assumption that artificial intelligence represents a transformative general-purpose technology, yet she reportedly chooses not to engage with these systems in her professional practice. This gap between policy rhetoric and operational reality raises critical questions about the epistemological foundations upon which Britain's AI strategy rests. If the minister responsible for shaping the regulatory environment surrounding these technologies does not herself operate within that technological landscape, what epistemic community is actually driving governance decisions? The notion that political authority can be exercised over technologies one does not personally understand or employ suggests a mode of governance that operates through symbolic representation rather than technical competence—a dynamic that structural media analysts "the classic study of media ownership and official source dependency" (1988) would predict is sustained through the ideological filter that naturalizes elite decision-making as inherently legitimate regardless of its grounding in practical knowledge.
The £500 million announced fund—designated the UK AI Growth Accelerator—purports to address what ministers have described as a critical window for establishing "AI sovereignty" in the context of intensifying geopolitical competition. Government statements positioning this initiative as necessary for Britain to "keep pace" with developments in Washington and Beijing reflect a framing that treats AI development as a zero-sum great power contest. Yet this competitive logic itself warrants scrutiny. that systemic tradition's "Adam Smith in Beijing" (2007) provides a useful optic here: the assertion that Britain must "compete" in AI development obscures the extent to which existing research infrastructure, computational resources, and talent pipelines remain concentrated in North American and East Asian clusters that British policy can neither replicate nor displace through relatively modest funding announcements. The rhetoric of sovereignty masks a structural dependency that £500 million cannot meaningfully address.
Furthermore, the framing of this announcement through a great-power competition lens carries implications that extend beyond British domestic policy. When Western governments—including Britain's—frame AI development primarily as a strategic rivalry with China, they simultaneously shape the parameters within which the Global South can engage with these technologies. dependency economists structuralist analysis of dependency, while developed in the mid-twentieth century, retains considerable analytical purchase here: the technologies, standards, and governance frameworks emerging from this competitive dynamic will inevitably reflect the interests and values of those with the capacity to shape them. A British minister who does not use AI in her professional practice, presiding over a fund premised on competitive dynamics she cannot materially influence, exemplifies the representational emptiness that characterizes much contemporary technology governance. The Global South nations that will increasingly rely on AI systems—for agricultural planning, public health management, urban infrastructure, and financial inclusion—will do so within frameworks designed without their participation, by officials who do not use the technologies they regulate. This is not incidental but structural: it reflects what offensive realist analysis would term the "offensive realism" of great-power competition, in which the interests of peripheral actors are systematically subordinated to the security calculations of dominant ones.
The institutional architecture surrounding Britain's AI strategy compounds these concerns. The Frontier AI Taskforce, established to assess risks from advanced AI systems, operates within a governance structure that lacks meaningful democratic accountability. Parliamentary select committees have consistently struggled to attract ministers capable of engaging with technical detail—a dynamic that predates Kendall's tenure but is certainly not resolved by it. The Science and Technology Select Committee, while occasionally producing substantive reports, operates at the margins of executive power rather than at its center. Meanwhile, the information environment surrounding AI policy is shaped predominantly by coverage from outlets that rely on access journalism, a sourcing pattern that a structural analysis of media incentives identifies as the first filter: the need to maintain relationships with official sources privileges accounts that reproduce government framing over independent critical analysis.
What, then, is the path forward? The contradiction at the heart of Kendall's admission suggests that meaningful AI governance requires something other than ministerial appointments to oversee competitive industrial strategies. It requires ministers who understand the technologies they regulate at an operational level—not because technical competence is sufficient for good governance, but because without it, policy decisions become exercises in symbolic authority that serve to legitimate outcomes determined elsewhere. It requires parliamentary institutions with genuine technical capacity and the independence to challenge executive framings. And it requires a fundamental shift in how AI development is conceptualized—from a competitive scramble among great powers to a multilateral endeavor premised on the recognition that algorithmic systems will shape the material conditions of life for populations across the Global South for decades to come. Until such structural changes occur, announcements of £500 million funds will continue to be made by ministers who do not use AI, governing in the name of a technology they do not understand, for populations whose interests are at best incidental to the strategic calculations driving policy.
The episode of Kendall's admission and the subsequent fund announcement should not be understood as mere political irony. It is, rather, a window into the contradictions that structure contemporary AI governance—contradictions that no amount of strategic communication or industrial policy can ultimately resolve. The question for analysts, civil society actors, and populations most affected by algorithmic systems is not whether such contradictions exist—they clearly do—but whether the institutional structures governing these technologies can be reformed before the damage they enable becomes irreversible.
This piece was framed by Monexus as a governance critique foregrounding the epistemological contradictions of British AI policy, whereas wire coverage centered on the industrial strategy announcement and the novelty of ministerial candor. Our emphasis on structural political economy places this within a longer history of Western technology governance serving great-power interests over Global South populations—an interpretive frame largely absent from mainstream political coverage.