Essay by Eric Worrall
Southampton Lecturer in Law Louise Du Toit is concerned that even EU regulators seem to be overlooking the AI Environmental Crisis.
AI laws overlook environmental damage – here’s what needs to change
Published: April 2, 2026 9.55pm AEDT
Louise Du Toit
Lecturer in Law, Southampton Law School, University of SouthamptonMore than 200 laws have been developed to regulate AI in more than 100 countries. Many of them focus on issues such as privacy, bias, disinformation, security and cybersecurity rather than the environmental consequences of AI.
AI is an energy-intensive and thirsty industry. It leads to huge greenhouse gas emissions, pollution and loss of nature. These impacts arise partly from the manufacture and use of energy-, carbon- and water-intensive “complex computer chips”, called graphics processing units (GPUs), for the training of AI models as well as increasing e-waste.
My research into the regulatory responses to AI in the EU and the UK highlights how laws often ignore the environmental implications of this big tech. The lack of stringent obligation in AI law and policy is concerning.
…
More transparency starts with AI developers having to disclose information about how much energy and water is consumed, how much carbon is emitted, the rare earth elements extracted and how much plastic is used during the AI production process.
Read more: https://theconversation.com/ai-laws-overlook-environmental-damage-heres-what-needs-to-change-279047
…
There’s a very simple reason even the EU is going easy on AI – they are terrified of falling behind.
Despite rather transparent attempts by tech companies to define AI as a climate saviour, the reality is AI is driving a tank through all their green pretensions.
Green energy and Net Zero, at least solar / wind / battery Net Zero, are utterly incompatible with AI. Either governments give AI a pass, while doing what is necessary to facilitate AI access to energy and cooling water with minimal disruption to ordinary people, or they accept being relegated to becoming economic has beens.
Something’s got to give and I don’t think it’s physics.
Physics just keeps doing what physics does.
And catches out the “settled science” folks every time.
Have a chuckle at this –
https://www.theweathernetwork.com/en/news/weather/severe/more-than-160-crashes-reported-by-calgary-police-sunday-amid-snowfall
Oh no!
Our green energy scam is incompatible with our artificial intelligence scam!
Whatever shall we do?!
They tried defining AI as a stepping stone to climate nirvana, but greens aren’t buying it
There is a natural synergy between future developments in AI and nuclear energy.
That would absolutely address any concern about CO2 emissions. But after decades of pushing useless renewables, it would be a bit embarrassing for greens to turn around and advocate nuclear as a way of containing AI emissions. So they avoid embarrassment by campaigning against AI.
To be fair, Eric, they don’t seem to be much a fan of NI, either.
Artificial Intelligence is no match for Natural Stupidity.
How to segregate distortion and misrepresentation from facts?
Apply logic.
Can You give an example of an existing industrial facility providing all of its own electricity? I know of none as it would be very expensive.
I am not sure this is what you are looking for, but the sugar industry uses cogeneration from burning bagasse, producing in-house electricity with some left over for sale, and steam for sugar processing.
Not an example. Wood industry does this too. Burning your waste is efficient but generating electricity without such waste is not.
Aughinish alumina in ireland.
most if not all oil refineries.
many many more.
because for energy intensive industrial facilities to depend on power utilities for their power would hugely extend the timelines for their development.
A facility needs say 100MW Power.(often more) It is far cheaper and quicker to do itself than to get that power included in the power utilities business plan, then into its capital expenditure program and then get that coordinated with the facilities engineering procurement and construction schedule. It would almost certainly cause a minimum of perhaps three years of delay to the facility commissioning.
ALL major energy intensive facilities I gave been involved in for over 40 years had their own generation.
Grid connections were for backup, and occasional supply to the grid.
Steel mills often operate on-site power plants, frequently using by-product gases (blast furnace/coke oven gas) or natural gas for cogeneration, though some historically rely on coal.
It is true AI uses lots of energy and water that is why we must insist AI spend its money producing energy and treating the water it uses. AI isn’t going away but the difference between AI and the useless renewables scam is that we can benefit from AI whereas renewables benefit from us.
How do data centers consume water? Where does the water go? Those claims are overblown.
I’m just glad these luddites weren’t around when the US built the Transcontinental Railroad. Can you imagine what they would have thought of a coal burning train belching thick black smoke and ash particles traveling over habitat-destroying railroads? And those trains used enormous amounts of water, as well. Now can you also imagine how the US would have expanded without those railroads transporting food from mid-west farms to heavily populated cities? Every city would need to be fed by nearby farms.
What I have gained combining my limited imagination with AI-provided research convinces me that this will be the AI revolution, and every bit as impactful as the Industrial Revolution. Fortunately, pursuing AI development will be a national security concern, and governments will override any effort to ‘derail’ it.
IMO, the left is leading the charge against AI because Big Tech abandoned the Climate Change bandwagon. The Tech Bros quickly realized they couldn’t embrace ‘green’ energy, ‘renewables’, Net Zero and get the power they need for their data centers.