We Thought We Were Building Services. It Turns Out We’re Renting the Ground Beneath Them - Governance and Digital Sovereignty 1
There’s a line from Yanis Varoufakis that refuses to go away. Not because it’s catchy, but because it’s uncomfortable:
Austerity for the many. Support for the few.
He was talking about banks, bailouts and the aftermath of 2008. Easy enough to file under macroeconomics and move on. But every so often, it resurfaces in places it wasn’t meant to reach. Local government is one of them.
Because if you squint slightly, what happened after 2008 doesn’t feel entirely remote. We tightened. We rationalised. We digitised. We were told to do more with less, and in fairness, we did. Quietly, persistently, often creatively.
But here’s the part we don’t always say out loud.
In becoming more digital, we didn’t just improve services. We changed the ground those services stand on.
---
I’ve been writing recently about agents. Not in the sci-fi sense, but the slow emergence of systems that act on our behalf. The idea that one day we’ll say, without irony, “I’ll have my agent speak to yours.” It sounds like a throwaway line. A bit of convenience wrapped in novelty.
But there’s a question underneath it that’s harder to ignore.
Where exactly are those agents meeting?
Because in local government, we already know the answer, even if we don’t phrase it that way.
They meet inside platforms.
---
Take any modern service in a council. Social care systems, finance, HR, customer platforms, planning, GIS. We don’t build these from scratch. We adopt, configure, integrate. It makes sense. It would be absurd to do otherwise.
But over time, something shifts.
The system stops being a tool and starts becoming an environment.
You don’t just use it. You operate within it.
Processes are shaped by it. Data is structured through it. Reporting is constrained by it. And when you want to change something, really change it, you discover where the edges are. Not always where you expected.
This isn’t failure. It’s design.
And it’s where Varoufakis’s more provocative idea comes in. His notion of techno-feudalism. The claim that we’re moving away from markets in the traditional sense, towards something closer to controlled digital estates. Not feudalism in the medieval sense, but in the way power is exercised. Access is granted. Activity is enabled. But the rules sit elsewhere.
You don’t have to agree with the label to recognise the shape of it.
---
In the PMO world, we spend a lot of time on risk. Red, amber, green. Mitigations. Owners. Review cycles. All necessary. All useful.
But the real work, the part that rarely makes it into the highlight report, sits one layer up.
Assumptions.
We assume the system will evolve with us.
We assume the contract gives us enough leverage.
We assume integration will remain manageable.
Most of the time, those assumptions hold. Until they don’t.
And when they don’t, the options are familiar. Work around it. Wait for a roadmap. Negotiate. Accept.
None of those are catastrophic. But they do reveal something.
Control, in a digital environment, is not binary. It’s negotiated.
---
I’ve been exploring this idea in different ways over the past year. The Prime Directive in Star Trek as a metaphor for intervention and restraint. Prompting AI as photography, editing as filmmaking. The question of whether it’s better to have a question without an answer than an answer without a question.
This feels like a continuation of that thread.
Because here, the question isn’t how do we deliver better services?
We’ve got quite good at that, within constraints.
The question is:
What kind of system are we delivering them in?
---
There’s a fair challenge to all of this.
We’re still operating in markets. Councils choose suppliers. Frameworks exist. Competition happens. Innovation is real. No one is forcing a local authority into a single platform at swordpoint.
All true.
But lived experience is rarely that clean.
Switching systems is expensive, disruptive, risky. Data models don’t translate neatly. Integrations unravel. Operational knowledge is tied to specific configurations. The longer a system is in place, the more it becomes part of the organisation itself.
Choice exists, but it narrows over time.
Not eliminated. Just… shaped.
---
And then, just as we’ve stabilised around that model, something else arrives.
AI, again. But this time not as a tool we dip into, but as something that begins to sit between us and the system.
A resident makes a request. An agent interprets it, navigates eligibility, completes forms, triggers workflows. On the other side, organisational agents coordinate resources, manage demand, optimise responses.
Efficient. Sensible. Probably inevitable.
But the earlier question comes back, sharper this time.
Where are those interactions taking place?
If they’re happening inside the same platform environments, and it’s hard to see why they wouldn’t, then the layer we interact with becomes further abstracted. Not just people using systems, but systems negotiating within systems.
At that point, “I’ll have my agent speak to yours” starts to sound less like convenience and more like protocol.
And protocol, in someone else’s environment, is never entirely neutral.
---
So where does that leave us?
Not with a grand rejection of platforms. That would be performative at best. Nor with a neat solution. This isn’t that kind of problem.
But perhaps with a shift in emphasis.
We keep doing the things we know matter. Good governance. Clear accountability. Thoughtful procurement. Strong delivery.
But alongside that, we get a bit more explicit about the ground we’re standing on.
We surface dependency as a strategic consideration, not just a technical detail.
We challenge assumptions about control, not just timelines and budgets.
We recognise that the environment we operate in is partly designed elsewhere.
And we ask, occasionally and deliberately:
Are we building services… or are we renting them?
Because if Varoufakis is even half right, that’s not just an economic question.
It’s a governance one.
Comments
Post a Comment