The Hidden Energy Costs of Artificial Intelligence
- AI’s energy use is rising rapidly, but poor transparency and fast-changing technology make accurate projections nearly impossible.
- Governments and companies are preparing for worst-case scenarios by accelerating energy development, sometimes at the expense of climate goals.
- The real environmental impact of AI is driven by large-scale industrial deployment, not individual user behavior or isolated queries.
Projections for the AI sector’s energy needs in the coming years are massive in scale. The International Energy Agency expects AI’s energy demand to double between now and 2030, presenting a serious challenge for energy security in many nations and regions where large data center developments are planned.
But planning ahead for data center development and their associated energy needs is an almost impossible task. The real and future energy use of artificial intelligence is incredibly hard to pin down due to the rapid growth and advancement of the technology, as well as the lack of disclosure requirements imposed on AI firms. Accurate projections for the future are therefore all but impossible, since we don’t even know how much energy AI is using in the present moment. But we do know that it’s a whole lot.
The result of all this uncertainty is an enormous amount of hand-wringing on the part of end-users and rapid - panicked, even - investment in new and expanded energy production capacity on the part of the private and public sectors alike. “The energy resources required to power this artificial-intelligence revolution are staggering, and the world’s biggest tech companies have made it a top priority to harness ever more of that energy, aiming to reshape our energy grids in the process,” stated the MIT Technology Review in a May 2025 report.
World leaders are left with little option but to prepare for the most intensive scenarios. Already, countries around the world are fast-tracking new energy development, often at the risk of climate goals. “From the deserts of the United Arab Emirates to the outskirts of Ireland’s capital, the energy demands of AI applications and training running through these centres are driving the surge of investment into fossil fuels,” Financial Times reported in August.
Better and more responsible policy around AI and its supporting industries will necessarily depend on better and more available data about AI’s energy use – but major questions remain unanswered about how much will that use fluctuate as levels of both integration and efficiency increase, and how much does the way that users interact with these platforms influence the energy footprint of large language models.
The subject of how much energy an individual AI query uses is currently a subject of much debate. There is even a question as to whether our politeness with large language models – using extra computing power to say please and thank you to models like ChatGPT – is directly driving up energy usage and costing companies like OpenAI millions. No matter how many times ChatGPT receives the input “thank you” it has to run a fresh “inference”, performing “a full computational pass through the model.” All those individual computations add up, and add up in a big way.
Of course, the concern over a few extra polite words is extremely small potatoes compared to the myriad other, more demanding ways that AI is being used. Nevertheless, a recent op-ed for The Conversation argues that the “persistence of the idea” that all those drops in the ocean add up to an important impact “suggests that many people already sense AI is not as immaterial as it appears.” The article goes on to state that “that instinct is worth taking seriously.”
On the other hand, this mentality is also diverting attention from the real problem of AI’s environmental impact. Individual queries and user-end activity is virtually negligible compared to what’s happening on the producer end of the equation. The spread of AI is not user-driven. Rather, it’s industry-driven, and being indiscriminately integrated across virtually every economic sector there is at a rapid pace, and with serious energy consequences.
“AI’s integration into almost everything from customer service calls to algorithmic “bosses” to warfare is fueling enormous demand,” the Washington Post reported last August. “Despite dramatic efficiency improvements, pouring those gains back into bigger, hungrier models powered by fossil fuels will create the energy monster we imagine.”
By Haley Zaremba for Oilprice.com
JP Morgan Boss Sounds Alarm on AI Bubble and Sticky Inflation Threat
- Jamie Dimon, CEO of JP Morgan, issued a stern warning to global markets, cautioning against underestimating geopolitical risks, sticky inflation, and the threat of an AI bubble.
- The bank announced its fourth-quarter 2025 earnings, which included a hefty increase in loan loss provisions to $4.7 billion, taking a seven per cent chunk out of the firm's profit.
- JP Morgan announced a £10 billion investment in the UK, including a new Canary Wharf headquarters, following Chancellor Rachel Reeves' attempts to foster closer ties with global financiers.
The boss of JP Morgan, Jamie Dimon, has warned global markets not to underestimate current risks, as the US banking giant hiked its provisions for bad loans.
The world’s most influential banker said that whilst the US economy had remained “resilient” and consumer and business trends were “generally healthy,” ongoing risks persisted.
“Markets seem to under-appreciate the potential hazards – including from complex geopolitical conditions, the risk of sticky inflation and elevated asset prices,” Dimon said.
Dimon also weighed in on growing agitation surrounding the independence of the Federal Reserve after Trump’s latest attack on chair Jerome Powell.
“Anything that chips away” at the central bank’s independence “is not a good idea,” he said on Tuesday, adding “everyone we know believes in Fed independence”.
The banking titan’s warning to markets come after he sounded the alarm on an AI bubble telling the BBC he was “far more worried than others.”
“Most people involved won’t do well. Some of the money being invested will probably be lost,” he said.
Fears of an AI bubble have sent jitters across the UK following pledges of capital injections from US giants into the UK economy.
Microsoft earmarked £22bn to build the country’s largest supercomputer and AI infrastructure. Meanwhile, chipmakers Nvidia and OpenAI laid out plans to create the largest AI computing facility in Europe.
Still, markets have continued to notch record highs despite ongoing concerns. The FTSE 100 smashed the 10,000 milestone in its first trading session of the year, whilst S&P and Dow Jones wrapped up last week’s with new highs after rallies in chipmakers.
The banking chief has also issued warnings around the rise of private credit, following the collapse of car parts maker First Brands and subprime auto lenders Primalend and Tricolour.
Dimon cautioned more “cockroaches” were likely to emerge in a credit downturn, telling an analyst call that some banks’ underwriting of loans to private credit “won’t be as good as you think”.
Dimon hikes provisions for sour loans
The renewed concerns came as JP Morgan released its fourth-quarter earnings update for 2025, where the bank notched $46.8bn (£35bn) in managed revenue.
The bank officially announced it will become the new issuer of the Apple Card, a move expected to bring over $20bn in card balances to the Chase platform.
But JP Morgan continued the trend of bulking up its financial cushion amid broader economic nerves, with loan loss provisions rising to $4.7bn in the final quarter from $2.6bn the year prior and $3.4bn in the third quarter.
The hefty increase took a chunk out of the firm’s bottom line, with profit falling to $13bn, a seven per cent drop compared to the prior-year quarter.
Chris Beauchamp, chief market analyst at IG, said: “This is another great set of numbers from JPMorgan, notable for strong client inflows and payments revenues… investors worried about overstretched equities can at least ease back on concerns about earnings.”
This week Dimon is expected to introduce Chancellor Rachel Reeves as they jointly host an event at next week’s World Economic Forum, according to Sky News.
The roundtable gathering in Davos, Switzerland, will be attended by bosses of the world’s biggest multinational firms and comes as Reeves attempts to curry favour with global financiers in her bid to make Britain an investment destination.
Reeves has maintained a close relationship with the American banking boss as part of her attempts to drive economic growth across the UK.
Following the Autumn Budget – where banks were able to skirt a highly-anticipated tax raid – JP Morgan announced a whopping £10bn investment into the UK with a new Canary Wharf office.
The project is expected to create an additional 7,800 jobs across construction and other local industries. Once finished, it will house up to 12,000 and serve as the bank’s main headquarters in the UK, and it will be the bank’s biggest presence across Europe, the Middle East and Africa.
JP Morgan contributes nearly £7.5bn to the local economy and supports 38,000 jobs.
By City AM
No comments:
Post a Comment