A deep dive into what Open Data means for water companies and regulation of the sector
Open Data in the water sector is in a bit of a conundrum right now. There needs to be enough data out in the open to prove its value, but you can’t prove its value until it’s out in the open. It’s a bit like Schrödinger’s cat, but for data. The good news is that it’s a challenge that’s starting to be solved.
Water regulation has always run on data. What’s changing now is who gets to see it, how fast, and what they’re allowed to do with it once they have it. That shift from data as a compliance artefact to data as a shared, usable public resource is what’s turning Open Data from a regulatory tick-box into one of the most important tools the sector has for building trust, driving innovation, and, increasingly, powering AI.
At Oakland Everything Data, we think Open Data is the future of water regulation. Here’s why.
What Ofwat requires from UK water companies
Ofwat has set clear expectations for UK water companies to open up their data by publishing priority datasets and a roadmap for how they’ll get there. After finding widespread public support for Open Data, but limited progress from the sector itself, the regulator has been direct about companies needing to take immediate action.
From ‘report to the regulator’, it’s now become ‘publish for anyone’, a genuinely new kind of ask for an industry that’s traditionally guarded its data closely. (Particularly given the sensitivity of infrastructure and customer information.)
And it’s not just an Ofwat initiative in isolation – the Open Data Institute has worked with the sector on a shared, industry-wide Open Data strategy, which is giving the push to publish some independent, cross-sector weight rather than a single regulator’s mandate.
How Yorkshire Water became Stream’s top-performing water company
When Ofwat introduced these requirements, Yorkshire Water (which serves over five million people) faced a first-of-its-kind challenge. There was no playbook, no established governance model, no shared understanding, even internally, of how to safely release data that had never been designed to leave the building.
Working alongside Yorkshire Water and the sector’s collaborative platform, Stream, we took a dual-track approach: the ‘do’ and the ‘design’.
- The ‘do’: On one track, we got moving straight away, delivering real Open Data releases via Stream to test the end-to-end process and prove, in practice, that data could be published safely.
- The ‘design’: On the other track, we built the long-term foundation: a full Open Data strategy and roadmap, a governance framework with clearly defined data owners, stewards and approvers, a value model to weigh benefit against risk, and the technical architecture to make it all repeatable.
The results
Rather than finishing a strategy document before doing anything, we did both at once to build momentum through visible wins while architecting something built to last.
The result speaks for itself: Yorkshire Water went from lagging behind on Open Data maturity to becoming the top-performing water company on the Stream platform. Compliant, yes, but leading, too.
And crucially, the process didn’t stop at what the regulator asked for. Yorkshire Water now has an engaged community of data owners and stewards, a growing pipeline of datasets in the works, and agreement in place to build an external Open Data hub of its own.
How these results look in practice
One of the clearest examples of what this looks like in practice? Reservoir levels. During periods of drought this year, reservoir data became newsworthy. And because it was already published on Stream, utilities didn’t have to field the same query over and over. They could just point people to the data. It meant less admin for the water companies, faster answers for the public, and more visibility for the shared platform itself. That’s Open Data doing exactly what it’s meant to do.
Yorkshire Water isn’t the only one moving. Northumbrian Water became the first UK water company to publish an Open Data strategy back in 2023, shortly after Ofwat’s initial ruling – a sign that this shift is genuinely sector-wide, not a single utility’s initiative.
Why format matters as much as access
But there’s a catch that trips a lot of organisations up: ‘technically available’ data doesn’t mean ‘genuinely usable’. A dataset buried in a PDF is ‘open’ (strictly speaking), but it’s not much use to a researcher, a journalist, or an app developer trying to build something with it.
Real ‘open’ data means machine-readable formats, proper metadata, and interoperability data that can be pulled straight into a spreadsheet or a tool like Power BI, rather than manually copied out of a document. It may sound like a small distinction, but it isn’t – and it’s what separates data that gets used and data that just gets published.
You can see what ‘done well’ looks like in the National Storm Overflow Hub, the world’s first, near real-time map of discharge data across nearly 14,000 storm overflows in England. The Hub:
- Lets swimmers and kayakers make informed decisions about entering the water
- Gives the public visibility into improvement plans for every single overflow
- Brings water companies together with independent oversight from the Environment Agency and Ofwat
All in all, it’s a good example of what happens when data is opened up properly rather than technically.
What Open Data enables for AI in water regulation
Generative AI models scrape unverified public web data by default, which means they rely on outdated or inaccurate data. By creating high quality standardised infrastructure with Open Data, we allow these machines to consistently find, access, and trust this information.
River Deep Mountain AI (RDMAI) has announced the open-source release of a suite of artificial intelligence and machine learning (AI/ML) models that it says are set to transform the way water quality data is collected and used. Right now, the utilities sector wants to use advanced machine learning and Large Language Models (LLMs) to:
- Predict critical infrastructure failures
- Track real-time catchment health
- Model climate resilience risks
But predictive AI can’t operate safely or accurately in a siloed environment. To find valid correlations, algorithms require continuous pools of standardised data.
None of this (the governance, the standards, the machine-readable formats) is really about AI on its own. But it’s the thing AI actually needs. You can’t build a reliable model on data nobody trusts, and you can’t scale a model across a sector that hasn’t agreed on what ‘clean’ even means. The work Stream and its member companies have done on Open Data is laying the groundwork for what comes next, whether it was designed this way or not.
Why Yorkshire Water started with what people actually wanted
If you think this happens because a regulator says so, then think again. It happens because someone in the organisation actually cares enough to make it work – and that’s arguably the harder part.
Part of what made the Yorkshire Water programme land was asking a different question from the start. Rather than publishing what a company assumes might be useful, it started with what people are actually asking for:
- Environmental Information Requests
- Customer queries
- Academic interest
These are all parts of the backlog worth working through first.
How a small shift in thinking paves the way for a big effect
In short, we need to release things that are actually useful to people first. This change in approach has a huge impact: it turns Open Data from something done to an organisation into something built with the people who’ll actually use it. It’s part of why the cultural shift matters as much as the technical one, and one that enables Open Data to be treated as a strategic asset in its own right.
Where Open Data and AI in water regulation is heading
So, if Open Data and AI adoption both continue on their current path, what does water regulation actually look like in five years?
In our opinion, water regulation will be far more insight-driven and even more focused on climate related initiatives in five years’ time. As Open Data matures and AI becomes embedded in regulatory workflows, the sector may move from manual reporting and human-dependent processes to more machine-readable data streams that artificial intelligence can validate and analyse in real time. This shift will reduce routine human intervention, improve data quality, and allow regulators and companies to focus on solving issues rather than chasing errors. It’ll also support risk detection, clearer accountability, and regulation that is based on live insight rather than retrospective reporting, which is something that the water industry is struggling with at the moment.
In June 2026, Ofwat published its first AI adoption plan for the water sector. For all of the details, plus our opinion on the implications it has on water companies, check out our blog: What Ofwat’s AI adoption plan actually asks of water companies.
Understanding the relationship between Open Data and AI
All this said, Ofwat and water companies need to recognise that AI is only as strong as the data it’s given. And this is exactly why we feel that Stream is so critical at Oakland Everything Data.
There’s a misconception that AI can simply ‘look at the whole internet’ and produce useful answers, but that often leads to responses that are inaccurate or irrelevant for water regulation. What actually unlocks value is accurate, clear, machine-readable Open Data published consistently – and this is where we get the real value from Stream.
When the industry publishes good data, AI can generate reliable insights which support better decision-making, which should lead to solutions grounded in trusted, accurate water data rather than generic web content.
The future of water regulation in an Open Data world
The foundations for Open Data are there. The standards exist, the governance models work, and, as Yorkshire Water shows, the results are measurable. The next chapter is about what’s built on top of water utilities opening their data.
For support with the transition to an Open Data water sector, please contact our experienced team of data specialists.
