The AI Action Plan and Britain’s Narrowing Path for Sustained Growth
The one-year anniversary of the UK’s AI Opportunities Action Plan has been met with a predictable flurry of reassurance. We are told that three-quarters of commitments have been delivered. Dashboards are green. The machinery of government is humming. If success were measured by activity, Britain would be leading the world.
But in the unforgiving arithmetic of the global economy, activity is not achievement. The government is pulling levers, yet those levers appear increasingly disconnected from the engine of the real economy. The comfortable fiction that Britain is “on track” obscures a harder truth: we are managing our own drift.
The relevant question is not whether the AI Opportunities Action Plan is being implemented. It is whether Britain still has the capacity to convert a general-purpose technology into national economic power before relative decline hardens into a new equilibrium.
The theatre of motion
Modern governance has become adept at the simulation of progress. Strategies proliferate. Metrics accumulate. Anniversaries are marked. None of this is trivial. But motion is not momentum.
Despite a year of intense policy activity, the diffusion of AI across the British economy remains shallow. We see pilots and experimentation, but not the deep, structural reconfiguration of firms and public services that underpins productivity growth. In a standard technology cycle, such caution might be prudent.
But AI does not follow a standard cycle. Its economic effects are shaped by increasing returns and coordination thresholds. Where early alignment occurs, benefits compound. Where coordination comes late, returns thin out and dependency grows. This produces economic hysteresis: firms that delay reorganisation do not simply catch up later. They enter a lower trajectory, increasingly dependent on intelligence designed, governed, and monetised elsewhere.
This is why dashboards mislead. They capture inputs – money allocated, programmes launched, reports published. They miss the only output that matters: are we altering Britain’s growth trajectory? At present, the evidence is weak.
The skills alibi
The current orthodoxy rests heavily on a comforting supply-side illusion, that if enough people are trained, growth will follow.
We have seen this before. We train, we upskill, we promote “literacy”. Yet productivity remains stubbornly flat. Why? Because skills are a derived demand.
British firms are not failing to adopt AI because they lack talent. They are failing because the risk-reward calculation is unfavourable. Regulatory ambiguity, unclear liability, weak assurance mechanisms, and fragmented data infrastructure make large-scale deployment appear reckless rather than rational.
To focus on skills while leaving these institutional constraints untouched is not policy; it is displacement. The burden of adjustment is shifted onto individuals while the structural incentives for organisational inertia remain intact. This is a familiar British failure mode. World-class human capital embedded in second-rate institutional architecture.
The absent state
This brings us to the uncomfortable core of the problem. Shifting a national economy onto an AI-driven trajectory requires a state capable of doing things Westminster has gradually unlearned. Not directing outcomes, but shaping markets.
That means moving beyond regulation as risk management towards market-making as a strategic function. It means using public procurement to accelerate learning and deployment, not merely to constrain behaviour. It means building assurance and audit capacity deep enough to make “safe deployment” an operational judgement rather than a rhetorical aspiration. Above all, it requires choice. Britain cannot lead everywhere. It must decide where it intends to be indispensable and accept the political costs that concentration entails.
This is not an argument for a larger state, but for a more disciplined one. In economies characterised by increasing returns and platform concentration, strategic neutrality is not prudence. It is acquiescence.
In earlier blogs for the Academy of Social Sciences, I argued that Britain’s central challenge lies less in ideas than in the erosion of institutional capacity to convert innovation into growth. AI is where that argument becomes unavoidable.
The narrowing path
Britain’s danger today is not sudden collapse. It is stabilisation at a lower level of economic ambition and influence. Drift is deceptive precisely because it feels manageable. It preserves the appearance of capability while underlying performance erodes.
The margin for error is narrower than it was a decade ago. Fiscal space is constrained. Productivity buffers are thin. Missed cycles cannot easily be absorbed. The anniversary of the AI Action Plan should therefore not be an occasion for self-congratulation. It should be a forcing moment for realism.
AI is not another policy domain to be managed incrementally. It is a structural inflection. Either Britain adapts its institutions to the coordination demands of this technology, or it accepts a future defined by dependence on systems designed and governed elsewhere.
The window remains open. But it is narrowing, and motion alone will not keep it open.
About the author
Professor Feng Li is Associate Dean for Research & Innovation and Chair of Information Management at Bayes Business School. His research investigates how digital technologies can be used to facilitate strategic innovation and organisational transformation across different sectors and domains. He advises senior business leaders and policy makers on how to manage the transition to new technologies, new business models, and new organisational designs in the digital economy.
Image: courtesy of Shutterstock