Gov Open Source: Conversation not Code
Government used to value open source.
The Technology Code of Practice stipulates: “Be open and use open source. Publish your code and use open source software to improve transparency, flexibility and accountability.”
But where have we ended up?
A department builds something new (all too often something that already exists) puts the code in a public GitHub repository, and walks away.
The code sits there. It becomes a digital artefact in a museum that nobody visits.
In fact when I “visited” one team’s code recently and messaged them about it they didn’t even realise their repo was publicly visible. “How did you find us?” I was their first ever visitor.
The code is open, but it’s not alive. It’s a monologue, not a conversation.
We have defined open source as a publishing mechanism, while simultaneously losing the most impactful element of this approach: Collaboration.
For many people they believe that open source requires effort without reward. But, when it’s a bureaucratic publishing standard that flows exclusively in one direction it’s no surprise that return on investment feels non-existent.
The frustration around what open source has become is palpable when you speak to open source advocates working in the system.
At the open source workshop I facilitated recently with the Government Digital Service the true nature of the problem was pretty clear to see. Rather than aligning with calls for a monolithic “government GitHub” that might only build a bigger silo, we uncovered a plea for simple discoverability.
The ability for one part of government to know what another part has already built and solved. To search for knowledge, for people, for ideas.
Because before we can collaborate, we have to connect. We need communication more than code.
The question of “how to discover” naturally leads to an even more challenging query – “where do we even look?”
It doesn’t take long to discover a fundamental vulnerability in government if you follow the white rabbit.
As things stand, government collaboration happens in proprietary, closed off spaces like Teams, Slack, and various WhatsApp groups (some approved, some less so). We don’t have a town square. Instead we are renting our virtual meeting spaces from vendors who can change the rules, raise the rent, or shut the doors at any time.
This mortgages our national technological sovereignty. We are building our future on someone else’s land.
The blueprint for such a town square already exists. It’s built on an open protocol called Matrix, with secure, enterprise ready versions offered by a UK based company, Element.
Importantly, this doesn’t have to be an all or nothing shift. Matrix is also designed to act as a bridge, connecting existing islands of Teams, Slack, and more communication capabilities. This allows government to retain optionality and escape vendor lock-in, without forcing an overnight exit.
Government has the opportunity to create a common space where all conversations can meet. A secure, sovereign foundation for government communication.
Once we have that foundation, the absurdity of our current development model becomes easier to see. The “Public Money, Public Code” mantra, as it exists today, ends with code going unloved in a lonely repo. But, when engineers can actually talk to each other, true open source development becomes possible in gov.
Instead of reinventing the wheel, with departments endlessly building their own complex and costly platforms, government engineers could be contributing to the upstream open source projects that everyone, including government, relies on.
This rapidly scales. Gov could be building on the shoulders of giants and reaping the rewards. A single contribution to a major open source project has an impact a thousand times greater than building another disposable, bespoke tool.
Working in this way would also make government a smarter customer that sweats its assets.
Rather than paying multiple vendors to solve the same platform challenge, government could consolidate its spending with a strategic partner like Red Hat. The immense purchasing power of government would act as a lever to influence the roadmaps of the key global projects that departments depend upon.
This would be transformative. It would turn a procurement budget for digital into a tool for global technological influence.
Government engineering talent would then be freed up to work on what truly matters. Differentiated, highly specialised, gov specific problems that will form the basis of our national services.
Why are we solving problems that have already been solved by open source communities?
Introducing an open source communication platform would be a provocative shift, but the impact of moving from a monologue to a conversation is far greater than many could imagine if all they know is the existing government open source approach.
This change has the opportunity to scale far beyond internal efficiency. It’s about our nation’s future relevance on a global stage.
While the UK is busy reinventing open source tools, the rest of the world is doing something very different.
China is investing heavily in the foundational AI technologies of tomorrow such as vLLM. France and Germany are building sovereign government chat and document platforms on open source. India built its digital identity system, Aadhaar, as an open source platform.
By becoming a major contributor to upstream open source the UK would gain global soft power as a custodian of critical digital infrastructure.
But the idea that we would miss out on having a say, and simply consume something we should have shaped is baffling. We cannot afford to miss this opportunity.
Especially when we can guide these upstream projects as a nation, and still have them be supported by enterprise open source companies. We can still outsource risk to vendors, even for products that we have co-developed.
The scale of this challenge can feel paralysing. It’s mammoth. But the solution isn’t a multi-year procurement programme that shifts the way that government purchases software.
Our starting point is simple, obvious, and can launch tomorrow.
Start with engineers. Treat our internal users, those expensive, brilliant engineers, like users.
To paraphrase Abi Noda, skilled engineers are the most expensive and valuable resource in any modern organisation. Why would we not want to make them as effective as possible?
So let’s start with a conversation.
This single act removes silos. It sets a new tone. It demonstrates a commitment to genuine, open collaboration.
And this collaboration will be the absolute bedrock for everything that comes next. Especially for AI where open source is more important than ever before.
Do we really want to keep launching our code into the void, or are we ready to collaborate?
It’s time for government to join the conversation.
