This isn’t about just buying shiny new software. This is about people. The real challenge, and the real opportunity, lies in public sector AI training. The government’s vision of algorithmic efficiency will remain a pipe dream unless it addresses the fundamental skills gap within its own ranks. The PAC’s findings are less of a gentle suggestion and more of a flashing red warning light on the dashboard of UK plc. It’s time to look under the bonnet.
The Land of Uneven Progress
So, where does bureaucratic AI adoption actually stand today? According to the PAC, the landscape is “uneven and at an early stage”. This is the polite, parliamentary way of saying it’s a bit of a mess. While some departments are running intriguing pilot schemes, others seem to be operating as if the internet is still a novelty. There’s no cohesive strategy, no unifying drive. It’s like giving one battalion in an army night-vision goggles and leaving the rest with candles. You don’t win a war that way.
The core of the problem, as the report points out, is a critical digital skills shortfall. We expect our civil servants to oversee, manage, and eventually deploy complex algorithmic systems, yet we’ve failed to provide them with the basic training to understand what these systems even do. How can a manager approve a project that uses machine learning if they can’t distinguish it from a spreadsheet macro? How can they guard against bias or ensure data privacy if the technical specifications read like a foreign language?
This isn’t a hypothetical problem. It’s the primary barrier preventing the government from tackling the monumental service backlogs that have plagued citizens. Without the know-how to implement AI effectively and ethically, we are simply storing up problems for the future. The conversation around AI in government has been dominated by what the tech can do, but the PAC rightly shifts the focus to what the people who will use it can do. And right now, the answer is: not enough.
Whitehall’s Best Kept Secret: The Operational Delivery Profession
Here’s where it gets interesting. Buried within the civil service is a potential solution, a sleeping giant. This is the Operational Delivery Profession (ODP). If you’ve never heard of it, you’re not alone. Sir Geoffrey Clifton-Brown MP, a member of the PAC, memorably described it as what “must be Whitehall’s best kept secret”. This isn’t some niche group; the ODP comprises around 290,000 people—over half of the entire civil service.
These are the people on the front lines. They process your taxes, issue your passports, and handle your benefits claims. They are the human face of the state. They are also, paradoxically, the very people who have been largely ignored in the conversation about government digital transformation. It’s a staggering oversight. You wouldn’t redesign a factory without talking to the people who work on the assembly line, yet that is precisely what has been happening.
Empowering the ODP is the single most important step towards successful bureaucratic AI adoption. These individuals understand the existing processes, the pain points, and the real-world needs of the public better than any consultant or Silicon Valley technology vendor. Training them isn’t just a matter of workforce reskilling; it’s about activating an unparalleled source of institutional knowledge. By turning the ODP into a digitally savvy, AI-literate workforce, the government can create a powerful engine for change from within.
Building the Human Engine for an AI Government
How do we move from acknowledging the problem to actually fixing it? The answer isn’t a single silver bullet, but a coordinated strategy focused on upskilling people and reforming processes.
From Apprehension to Aptitude: Upskilling the Civil Service
The goal of public sector AI training should not be to turn every civil servant into a data scientist. That’s a fool’s errand. The goal is to cultivate AI literacy. It’s about giving them the confidence to ask the right questions.
Think of it like learning to drive. Most of us don’t know how to build a car engine, but we know what the accelerator does, when to use the brakes, and how to read the warning lights on the dashboard. We need to equip civil servants with the same level of operational intelligence for AI. This means training that covers:
– The Basics: What is machine learning? What are its limitations?
– Ethical Guardrails: How do we spot and mitigate algorithmic bias? What are the data privacy implications?
– Practical Application: How can AI solve a specific backlog in my department? What would a successful implementation look like?
This is the foundation of any meaningful workforce reskilling programme. Without it, any investment in AI technology is built on sand.
Making the ODP a Destination
For too long, a role in operational delivery has been seen as a starting point, not a career. To change this, the government must actively promote the ODP. This means creating clear pathways for advancement, celebrating innovation from the front lines, and ensuring that the profession is seen as vital to the future of public service.
Sir Geoffrey Clifton-Brown’s comment about the ODP being a “secret” should be a source of embarrassment for Whitehall. The Cabinet Office and departmental leaders need to launch an internal and external campaign to raise its profile. When the people doing the work feel valued and see a future, they are more likely to embrace and drive change.
Supercharging the Surge Team
The government does have one promising tool in its arsenal: the Surge and Rapid Response Team (SRRT). This flexible group of civil servants is deployed to departments facing acute service pressures. The PAC report notes that the SRRT supported 75 deployments in the past year, a testament to its utility.
But its potential is so much greater. Instead of just being a team of firefighters sent to douse the latest administrative inferno, the SRRT could be repurposed as a transformation catalyst. Imagine if every SRRT deployment also included a mandate to identify opportunities for process automation and AI, and to leave behind a small, trained team to carry the work forward. It could become a travelling academy, spreading best practices for public sector AI training and implementation across government.
Fixing the System, Not Just the Symptoms
Of course, even the best-trained workforce will struggle if the underlying systems are broken. The huge service backlogs aren’t just a symptom of a skills shortage; they are a symptom of decades of underinvestment and creaking, inefficient processes. Plastering AI over a bad process just creates a faster bad process.
A true government digital transformation requires a willingness to engage in the unglamorous work of process re-engineering. It means asking fundamental questions about why things are done a certain way and being brave enough to scrap what doesn’t work. For example, before you build an AI to scan passport applications, perhaps you should first simplify the application itself.
The PAC’s recommendations, detailed in their report on the challenges in implementing digital change, provide a clear roadmap. The government needs to:
– Commit to a Sustained Investment: A long-term, properly funded programme for workforce reskilling is non-negotiable.
– Elevate the ODP: Appoint a visible, empowered leader for the profession who can champion its development.
– Integrate Skills and Technology: Ensure that every digital project has a parallel training and skills component built in from the start.
The Path Forward
The convergence of algorithmic power and public service is inevitable. The question is whether the UK government will manage this transition proactively or be dragged into it by crisis. The findings published by institutions like the Public Accounts Committee are not just another report to be filed away; they are a call to action.
The future of efficient, fair, and effective public services depends on the choices made today. Investing in technology is easy. Investing in people is hard, but it’s the only thing that will deliver a genuine transformation. By unlocking the potential of the 290,000-strong Operational Delivery Profession, the government has a historic opportunity to build a civil service that is not just fit for purpose, but fit for the future.
The journey ahead is complex, but the first step is simple: stop treating the people on the front lines as an afterthought. They aren’t the barrier to a digital government; they are the key to building one.
What do you believe is the single greatest obstacle to making this kind of deep, systemic change a reality in the public sector? Share your thoughts below.
