Since gaining power in 2024, Keir Starmer’s Labour Party have announced multiple AI data centre projects. What they haven’t announced is the potential impact these AI data centres will have on the UK’s water infrastructure.

There’s a very simple reason why they haven’t announced this, and it’s because they don’t know.

As we’ll explain, the government’s “systemic planning failure” has opened us up to staggering risks. They’ve also created a situation in which they may not be able to deliver key manifesto commitments on growth or home building.

To make matters worse, it’s not even certain there’ll be any pay off for ordinary Britons from these AI data centres.

Failure over AI data centres

The Government Digital Sustainability Alliance (GDSA) was established to provide guidance on how we can achieve “digital transformation” without putting sustainability targets at risk. In aid of this, they published a report titled Water use in AI and Data Centres.

As the report highlights, we were already facing significant water scarcity, and AI data centres will only worsen this issue. What we don’t know is how much worse they’ll make it, because tech companies don’t have to report, monitor, or estimate how much water they use.

We’ll repeat this, because it’s actually a little difficult to believe when you first hear it: AI data centres do not have to report, monitor, or estimate how much water they use.

What this means is that whenever you hear the government talking about water targets, those figures are void, because they do not account for a gargantuan drain on the system which will only grow over the coming decades.

Despite the gravity of the report, the only mainstream coverage we could find was a paywalled article in the Telegraphtitled Starmer’s AI drive ‘will lead to hosepipe bans’. We’re unsure why this report has received so little attention, but we aim to rectify that here.

“Critical oversight”

On September 16, the government announced several new data centre projects, including:

• A £22bn investment from Microsoft which will include data centres supporting cloud and AI infrastructure.
• A Google data centre in Waltham Cross.
• £1.5bn from CoreWeave going towards AI data centre capacity.

When we first read about this, we set about looking for the impact assessments which the government must surely have carried out to confirm these data centres wouldn’t completely undo our environmental targets – that they wouldn’t place an intolerable burden on water availability or drastically increase energy bills (more on this in a coming article).

We could not have been more wrong.

This is how the GDSA describe the situation (emphasis added):

A critical oversight in current UK water resource planning is that the measures proposed by water companies to meet the projected 2050 deficit do not account for the water needs of novel infrastructure or data centres. This represents a systemic planning failure, indicating a disconnect between the nation’s digital economy growth ambitions and its fundamental resource management strategies.

They also explain that this is one colossal problem on top of another (emphasis added):

The UK already faces a projected daily water deficit of nearly 5 billion litres by 2050, a challenge exacerbated by climate change and population growth. A critical policy gap exists as current national water resource plans, including those finalised in 2025 by water companies, do not adequately account for the burgeoning demand from infrastructure such as AI data centres. This oversight risks intensifying water stress in already vulnerable regions, potentially leading to social and environmental conflicts and hindering economic development.

As the GDSA have stated, we have no idea what impact AI data centres will have, because we’re not forcing these companies to report on usage. We do have some idea, however, as the GDSA highlights:

AI is predicted to lead to an increase in global water usage from 1.1bn to 6.6bn cubic metres by 2027. This is equivalent to more than half of the UK’s total water usage.

This increase will not be experienced evenly across the globe. The UK has the third highest amount of data centres, and we’re aiming to have similar growth in the area of AI data centres. As such, we’ll experience a higher drain on resources than other countries, even as the profits primarily benefit US tech companies:

AI data centres: water guzzlers

We’ve not explained this yet, but ‘data centres’ (DCs) and ‘AI data centres’ (AIDCs) are not the same thing. The vast majority of data centres in the UK are DCs, with AIDCs set to start coming online in the next few years. AIDCs place a significantly higher burden on electricity and water needs, as this video explains:

This isn’t to say that DCs don’t place an unbearable strain on water or electricity, as a recent BBC report found:

Data centres powering artificial intelligence (AI) in Scotland are using enough tap water to fill 27 million half-litre bottles a year, according to data obtained by BBC News.

AI systems such as the large language models (LLMs) that power OpenAI’s ChatGPT and Google’s Gemini require warehouses full of specialist computers.

The equipment is power-hungry, consuming large amounts of energy, but they also use tonnes of water in their cooling systems to stop the servers overheating.

Freedom of Information data shows the volume of tap water used by Scotland’s data centres has quadrupled since 2021.

Usage has already quadrupled, with many more DCs and AIDCs set to come online in the next few years – data centres which will not have to monitor or report on water usage.

The value of AI

By now, you’ve likely heard we’re in an ‘AI bubble’. The reason we’re in a bubble is because no one has managed to generate sustainable profits from AI, even though investors have pumped hundreds of billions into the sector. As tech journalist Ed Zitron wrote back in August:

An MIT study found that 95% of organizations are getting “zero return” from generative AI, seemingly every major outlet is now writing an “are we in a bubble?” story, and now Meta has frozen AI hiring. Things are looking bleak for the AI bubble, and people are getting excited that this loathsome, wasteful era might be coming to an end.

This isn’t to say there will be no value from the current wave of AI. AI tools are being used to accelerate medical breakthroughs, for example, but the ‘hyperscaling’ we’re witnessing isn’t curing cancer; it’s facilitating industrial scale ‘slop’ like the following:

Everyone: “Sora 2 is so amazing! You can’t even tell it’s ai”

Me: Alright, let’s see how realistic this is. “Sora 2, make a video of me high fiving an elderly woman and her falling into a lake”

Sora 2: pic.twitter.com/wDSn6reCwt

— GJake (@GernaderJake) October 17, 2025

You can tell these AI companies are getting desperate to show a return on investment, because OpenAI’s Sam Altman recently announced that ChatGPT would allow users to generate erotica. At the same time, he promised ChatGPT would soon be able “mitigate serious mental health issues”:

We made ChatGPT pretty restrictive to make sure we were being careful with mental health issues. We realize this made it less useful/enjoyable to many users who had no mental health problems, but given the seriousness of the issue we wanted to get this right.

Now that we have…

— Sam Altman (@sama) October 14, 2025

Altman had to make these reassurances because of a growing phenomenon called ‘AI psychosis‘ in which AI text generators amplify the delusions of users. This phenomenon may already be claiming lives:

When a teenage boy told ChatGPT about his suicide plans, it said: “I won’t try to talk you out of your feelings.” His family is suing after ChatGPT actively discouraged him from seeking help, offered to help him write a suicide note and even advised him on his noose setup. pic.twitter.com/pRijpY1jYV

— Angela Yang (@Angela_Y_Yang) August 26, 2025

These AI barons keep rushing their products out without confirming they’re safe, and politicians like Keir Starmer are helping them.

So much for the supposed aims of the Online Safety Act, eh?

So much for pretending we’re tackling climate change.

And if you’re wondering why the government is treating data centres as if they’re equally as important as water or energy, it’s because legally speaking they are. Since 2024, we’ve classed them as ‘Critical National Infrastructure’, and as we explain here, the designation means local areas are required to provide Microsoft and Google with all the water they can squander.

This is AI psychosis at the national level.

Government response

We wrote to the Cabinet Office to ask the PM the following:

• Does the Prime Minister have any comment on the fact that this failure to properly regulate US tech companies essentially means the UK is now flying blind on the issue of water availability?
• Additionally, does the Prime Minister have any comment on the fact that home building and growth targets may be impossible to achieve as a result of the knock-on effects of data centre-driven water scarcity?

These are quite clearly top-level questions which should be answered by the PM. Starmer didn’t answer, however, with his spokesperson instead directing us to the Department for Science, Innovation and Technology (DSIT). Perhaps unsurprisingly, the DSIT didn’t reply either, and neither did the Department for Environment, Food & Rural Affairs (DEFRA). ‘Unsurprising’ because what could they say, really?

If they didn’t know it looked bad, it wouldn’t all be happening in relative obscurity.

Featured image via the Canary

By Willem Moore

From Canary via This RSS Feed.