It was as long ago as 1982, back in the pre-privatisation days of the Central Electricity Generating Board, that the idea of building a new nuclear power plant in Suffolk – Sizewell C – was first mooted.
At that time, construction had yet to begin on the neighbouring Sizewell B, which for now remains the youngest of Britain’s operating nuclear power plants.
The first planning application was filed as long ago as 1989 and there have been countless false starts since.
The theoretical cost of construction was pushed up when Margaret Thatcher‘s government insisted that any company building a new nuclear power station would also have to have funding in place for not only its construction but also for the disposal of waste and the eventual decommissioning of the plant.
That proved a major obstacle to new nuclear build which was then further held up by Tony Blair’s reluctance to take on opponents of new nuclear build in his own party – although, in 2006, he eventually committed to the cause, as did his successor, Gordon Brown.
Hinkley Point C, the UK’s first new nuclear power station in a generation, was the upshot.
New financing key to unlocking nuclear
More on Energy
Related Topics:
Yet the construction of the Somerset plant is years behind schedule. EDF, the French energy giant building it and which will construct Sizewell C, originally envisaged it opening in 2017. Hinkley Point C is also billions of pounds over budget.
And the coalition government’s decision to guarantee EDF a fixed price for the energy generated at Hinkley Point C, which was necessary to persuade the French company to go ahead with the project, was subsequently heavily criticised.
Advertisement
The National Audit Office (NAO) said the agreement had locked consumers into a “risky and expensive” project – although, ironically, the deal now looks good value following this year’s spike in wholesale electricity prices.
The NAO’s report did, though, make subsequent governments wary, once more, of new nuclear build.
Theresa Mayimmediately demanded a review of Hinkley Point C on becoming prime minister and, even though her government ultimately approved the project, she also took note of a suggestion in the NAO’s report that new funding models be considered for subsequent new nuclear power stations.
That, in a nutshell, is why it has taken so long for Sizewell C to finally get off the ground. These plants are so monstrously expensive to build that no private sector company is willing to bear all of the risks themselves without some support from government. It is also why the likes of Japan’s Hitachi and South Korea’s Kepco have reluctantly walked away from building new nuclear plants at Wylfa on Anglesey, Oldbury in Gloucestershire and Moorside in Cumbria.
So key to unlocking the project has been coming up with a new way of financing it.
The solution
The government’s solution is the funding model known as Regulated Asset Base (RAB) – the means by which other major infrastructure projects, such as the £4.3bn Terminal 5 at Heathrow Airport, have been financed.
Under this arrangement, rather than guarantee whoever builds Sizewell C a set price for the electricity it generates, taxpayers will be taking risk alongside other investors.
This is why the government is investing an initial £700m in the construction of the plant although, with the total cost likely to come in at between £20-£30bn, that will only go so far.
The other elements in the RAB model include electricity consumers – households and businesses – paying for the plant while it is still under construction through their bills.
This is how, for example, the £4.13bn Thames Tideway tunnel now under construction is being financed. A share of the cost of the project, which is aimed at preventing sewage spills into the Thames estuary as well as future-proofing London’s sewerage system for expected population growth, is being met by customers of Thames Water on their bills.
The arrangement means taxpayers share in the pain of any cost-overruns. Other crucial aspects of the RAB model include an ‘economic regulatory regime’ (ERR), overseen by an independent regulator, who determines the extent to which investors and taxpayers will share the risks by setting the amount of revenue that EDF will be allowed as it builds Sizewell C.
Unknown sums but less risk
The government has yet to make clear the sum that billpayers will have to contribute towards the new power station but newspaper reports have suggested it will be in the region of an additional £1 per month per customer.
The Department for Business, Energy and Industrial Strategy said today that the lower cost of financing a large-scale nuclear project through this scheme was “expected to lead to savings for consumers of at least £30bn on each project throughout its lifetime” compared with the existing arrangements governing the financing of Hinkley Point C.
Image: Big Carl, the world’s biggest crane, in action at Hinkley Point C nuclear power plant near Bridgwater in Somerset
So in theory, while there is a risk attached to building Sizewell C, the funding model proposed appears to be less risky than the way in which Hinkley Point C has been financed. The ultimate cost to electricity consumers in the latter case was dictated simply by a decision made a decade ago on the price that EDF would be promised for its power. It currently looks good value but, for much of the last decade, it has not.
Yet the RAB model does have its critics.
Less incentive to control costs
Steve Thomas, emeritus professor of energy at the University of Greenwich, has argued that, by removing construction risk from EDF, the company has less of an incentive to control construction costs. With Hinkley Point C, EDF has had to bear the cost of any over-runs. With Sizewell C, taxpayers would be on the hook.
Professor Thomas argues that this is particularly worrying because he believes EDF’s cost estimates are too optimistic. He has also argued that the £1-a-month levy on household bills, should it come to pass, is also potentially flawed because of assumptions it is making about borrowing costs.
Less risky, for now, appears to be the ownership of Sizewell C. Objections to the involvement of the Chinese state-owned company China General Nuclear, originally raised by the May government, have resulted in the company now being bought out of its interest in Sizewell C. The project will instead be jointly owned by EDF and the UK government – although there has been speculation that new investment could also be brought in from the sovereign wealth fund of the United Arab Emirates.
There are, though, some other objections. The idea of building small modular reactors by companies like Rolls-Royce has won support on the basis that the technology could be cheaper and more scalable than big projects like Sizewell C. They would also, in theory, involve less cost in adapting the national grid.
Image: Prime Minister Boris Johnson during a visit to EDF’s Sizewell B nuclear power station in Suffolk.
The EDF question
Another risk concerns EDF itself. The company recently had to be bailed out and fully nationalised by the French government following the spike in wholesale prices.
But this means EDF is now effectively run at the behest of the French government. France is also anxious to build new nuclear power plants. Should EDF become cost-constrained it is perfectly plausible that the French state would direct it to focus on its domestic projects rather than its ones overseas.
There have already been hints of this.
EDF’s former chairman and chief executive Jean-Bernard Levy, who was effectively fired by President Macron after opposing nationalisation, was a strong supporter of Sizewell C but was hampered by the French government’s constant demands for more information on the project.
One final risk is that electricity demand does not increase in the way that the government is assuming and that Sizewell C’s output may not be needed.
However, with electricity demand projected to double as the UK decarbonises, that feels less worrisome than some other factors – and particularly now Vladimir Putin’s war on Ukraine has highlighted the importance of the UK having more indigenous sources of energy.
Lloyds Banking Group has set aside a further £800m to cover estimated costs associated with the car finance mis-selling scandal.
The bank said the sum took its total provision to £1.95bn.
It had been assessing the impact since the Financial Conduct Authority (FCA) revealed last week it was consulting on a compensation scheme, with up to 14.2 million car finance agreements potentially eligible for payouts.
The regulator had previously found that many lenders failed to disclose commission paid to brokers, which could have led to customers paying more than they should have between April 2007 and November 2024.
Eligible customers could receive an average of £700 each under the proposals.
Lloyds said on Monday that it would be contributing to the consultation to argue a number of points.
More from Money
It said: “The Group remains committed to ensuring customers receive appropriate redress where they suffered loss, however the Group does not believe that the proposed redress methodology outlined in the consultation document reflects the actual loss to the customer. Nor does it meet the objective of ensuring that consumers are compensated proportionately and reasonably where harm has been demonstrated.
“In addition, the approach to unfairness in the redress scheme does not align with the legal clarity provided by the recent Supreme Court judgment in Johnson, in which unfairness was assessed on a fact specific basis and against a non-exhaustive list of multiple factors. The Group will make representations to the FCA accordingly.”
Please use Chrome browser for a more accessible video player
0:58
Car finance: ‘Don’t use a claims firm – here’s why’
Shares in Lloyds, which fell last week when the bank warned of a potential “material” increase in its provisions, gained more than 0.5% on Monday.
The estimated compensation figure came in below the sum some financial analysts had predicted.
The shares remain more 50% up in the year to date.
Another listed lender exposed to car loan mis-selling is also expected to raise the amount it has set aside.
Close Brothers, which has a £165m provision currently, saw its shares tumble 7% when it admitted an increase was likely once its analysis of the compensation consultation documents was completed.
Car finance makes up approximately a quarter of its total loan book.
The budget may still be more than six weeks away, but rumours of U-turns and changes are already in full swing.
Over the last few days, there have been multiple reports that those inside Whitehall are considering tweaks to the controversial inheritance tax (IHT) reforms on farms announced this time last year.
Plans to introduce a 20% tax on estates worth more than £1m drew tens of thousands to protest in London, many fearing huge tax bills that would force small farms to sell up for good.
Now there are reports the tax threshold could be increased from £1m to £5m (£10m for a married couple) – a shift that would remove smaller farms from being liable to pay.
Please use Chrome browser for a more accessible video player
0:43
From February: Farmers continue tax protest
Senior figures in farming have long believed a rise could be the solution to save the smaller farms and it would satisfy most.
However under the proposals, the 50% relief on IHT would be removed for farms above the new threshold.
That means bigger farms, responsible for producing a large amount of produce in our supermarkets, could bear the brunt of the tax burden with the Treasury potentially increasing revenues.
More on Farming
Related Topics:
Two senior farming figures told me today that while a threshold increase is welcome, it does nothing to solve an “insolvable” problem.
Big farms have more land to sell, but then they become smaller farms and either produce less, or even divide up, to avoid the tax entirely.
Richard Cornock runs a small dairy farm in south Gloucestershire, which has been in his family since 1822.
Image: Richard Cornock plans to pass his farm on to his son
He hopes to pass it on to his son Harry, who is now 14 and training to become a farm manager.
“I’ve been under so much stress like most farmers worrying about this tax,” he said. “And I really hope they do push the boundaries on the thresholds, because the million pounds they propose at the moment is ridiculous.
“It’s been on my mind the whole time to be honest. I even looked into getting life insurance to insure my life and I can’t get it because I had a heart condition. And that was one way I thought I might be able to cover my kids…”
We paused our chat as he was too upset to continue – an illustration of the stress farmers like him have been under over the last 12 months.
Image: Tens of thousands from the farming community took part in protests in London. Pic: Reuters
The government says it won’t comment on “speculation” about any possible changes, but it has previously defended the IHT reform, saying most estates would not pay and that those who will be liable can spread payments over a decade.
Labour is under pressure to do something to appease the angry farmers, a rural vote that turned from the Conservatives at the last election.
I ask Richard whether any tweak or row back on IHT will restore faith in Labour?
The 38-year-old writer lost 70% of his clients to chatbots in two years.
His is one of 40 job roles that AI is fast replacing, according to conversations the Money team had with industry experts, researchers, and affected workers.
“It’s a betrayal,” says Turner, who earned six figures as a freelancer before the rise of generative AI.
“You’ve put your heart and soul into it for so long, and then you get replaced by a machine.”
He adds: “You always think ‘it’s never going to happen to me’.”
Image: Joe Turner
Around 85% of the tasks involved in Turner’s job could be performed by AI, according to research published by Microsoft in July that has gone largely unnoticed.
The tech giant’s analysis of 200,000 conversations with its Co-Pilot chatbot concluded it could complete at least 90% of the work carried out by historians and coders, 80% of salespeople and journalists, and 75% of DJs and data scientists.
Also in the top 40 most exposed jobs were customer service assistants (72%), financial advisers (69%) and product promoters (62%). Search the table below to see how your role fares…
Speaking to the Money team, senior Microsoft researcher Kiran Tomlinson insists the study “explores which job categories can productively use AI chatbots, not take away or replace jobs”.
Turner for one doesn’t buy this. “That’s what they want to market it as,” he says.
Experts we spoke with were just as sceptical of Microsoft’s optimism.
“If you were to look at these jobs in three to five years, there’s a very good chance they’ve been replaced entirely,” says an AI consultant with more than a decade of experience deploying the tech in nearly 40 companies.
“Except in areas where they are either relationship-driven or very judgmental,” they add, speaking on condition of anonymity due to their commercial relationships with a range of SMEs, multibillion-pound funds and public bodies.
“These types of jobs are by nature most likely to be replaced entirely by the tool,” agrees AI researcher Xinrong Zhu, an assistant professor at Imperial College London.
“We’re living in a world where we’re witnessing a very important turning point.”
Image: Xinrong Zhu
It’s a verdict echoing job cuts announced by major companies over the summer.
Buy now, pay later firm Klarna shrunk its headcount by 40% due to investments in AI and a hiring freeze, while boasting its chatbot was doing the work of 700 employees.
Microsoft itself said it was laying off 15,000 employees while investing £69bn in data centres to train AI models and reportedly using AI to save $500m in its call centres.
Amazon chief executive Andy Jassy said he expected to “reduce our total corporate workforce as we get efficiency gains from using AI extensively”.
But don’t take this at face value, says the AI consultant. Just because AI will take jobs doesn’t mean it can right now: “I wouldn’t say AI is in a position that you can then generate layoffs immediately: What you tend to see in most businesses is hiring freezes.”
The UK hasn’t had a sharp decline in postings for the jobs most threatened by AI, but they grew four times slower than the least threatened jobs between 2019 and 2024, according to PwC’s AI jobs barometer.
“AI is being used as an excuse,” the consultant says.
“There’s a load of macroeconomic effects that are actually causing [job cuts].”
It’s the Money blog’s usual suspects: Increases to employer national insurance, the cost of hiring and the cost of energy – not an AI takeover.
But, they say, “that’s not to say it won’t happen next year.”
Some 78% of global businesses anticipate increasing their overall AI spending this fiscal year, a Deloitte survey found.
Approximately 40% of employers expect to reduce their workforce where AI can automate tasks, according to a World Economic Forum survey.
An email that changed everything
Freelancers may, then, be the canary in the coal mine.
Demand for gigs related to writing and coding fell by 21% within eight months of the release of ChatGPT, according to a study conducted last year by Zhu.
“The magnitude really surprised us,” she says.
It wouldn’t have surprised Turner.
A few months earlier, in December 2023, he received an email from a website where he’d worked for a decade.
“Do you ever use AI?” it read. “No,” he replied.
That was the last time he heard from them. Overnight, £30,000 was wiped from his annual income.
“I went on their website and I realised they had started using AI instead of me,” he says.
One by one, most of his other clients followed suit.
“It was just a complete desert,” he says of the job landscape.
If you listen to the heads of some leading AI companies, you’d be forgiven for thinking this desert is just one apocalyptic vista at the end of the working world as we know it.
Dario Amodei, chief executive of Anthropic, has warned AI could “wipe out half of all entry-level white-collar jobs”, while OpenAI boss Sam Altman said entire job categories would be “totally, totally gone”.
“They want to glorify the models,” says Dr Fabian Stephany, a Labour economist at the University of Oxford and fellow at Microsoft’s independent AI Economy Institute.
Impersonating a big tech boss, he continues: “‘Oh wow, look, if we can automate away 50,000 people, then that technology must be really tremendous – so you should be investing in our company!’
“I would advocate to have a bit of more of a cooled down, pragmatic approach.
“Think about it as a technology and look at how technology has been interacting with the labour market in the past.”
Image: Fabian Stephany
Inventions that revolutionised the workplace
Take Richard Arkwright’s invention of the Spinning Jenny in 1769, which churned out huge quantities of yarn to make cloth in some of the first factories at the start of the industrial revolution.
While putting home weavers out of a job, it increased the need for mill workers hundreds of times over, says Stephany.
Henry Ford’s invention of the assembly line in 1913 had a similar impact when it reduced the time taken to make a car from 12.5 hours to 1.5 hours.
Speed lowered production costs and forecourt prices, increasing demand, sales and the number of staff hired to fulfil them.
For the same reason, the invention of the ATM in 1967 led to more bank teller jobs despite automating one of their key functions – something Microsoft was keen to point out.
“Our research shows that AI supports many tasks, particularly those involving research, writing and communication, but does not indicate it can fully perform any single occupation,” Microsoft’s Tomlinson says.
Indeed, the study shows 40 jobs where AI can perform just 10% or fewer tasks.
Tradespeople feature heavily, like painter-decorators (4%), cleaners (3%) and roofers (2%).
Surgical assistants (3%), ship engineers (5%) and nursing assistants (7%) also make the list.
But history also includes a list of the losers of technological innovation.
Replacing horses with tractors wiped 3.4 billion man hours from American farmwork annually by 1960, according to research by economic historian Professor Alan Olmstead.
Spare a thought, too, for the pinsetters once responsible for stacking bowling alleys, who were more or less eliminated by the Automatic Pinspotter unveiled in 1946.
Quantity does not mean quality, either: Arkwright’s millers faced exhausting and repetitive 13-hour shifts in extreme noise, heat and dust.
How fulfilling would working with an AI be?
“Sterile and just not interesting, uniform and bleak and surface-level and hollow” is how Turner described its work after trying AI at the request of a client.
“Cars were a solution – a car was a horse that never got tired. But if you look at AI the same way, it’s basically saying: ‘There aren’t enough rubbish books out there, we need to make more.'”
More human work, not less?
That’s not what it’s for, though, says the AI consultant.
“I don’t see an AI right now coming up with wonderful ideas for creative writing authors,” they say.
But what it’s good at is taking an author’s idea and making a first draft extremely quickly, they explain.
“Now, does that mean we have fewer authors or does that mean we have more?”
The consultant’s optimism comes from seeing AI create extra human work at some of the companies that hired them.
A landscaping firm used ChatGPT to generate personalised services to upsell to existing customers.
At a pension provider with 350,000 scheme members, AI saved “literally thousands of hours” by scanning millions of notes, PDF documents and email chains for spousal support agreements.
That might seem like work stolen from a law firm at first glance, but it likely wouldn’t have been undertaken at all without AI due to the extreme cost of manual labour, says the consultant.
The cost of starting a digital business has also shrunk dramatically, he adds, if you use AI to organise a website, workflow, marketing and employment contracts.
“You end up in a world where you could have thousands more small start-ups because the cost of failure is so much lower.”
Image: Pic: iStock
The ‘losers of technological change’
Such a positive attitude would do little to convince veteran audio producer Christian Allen, who has lost gigs worth £7,000 to AI in the past year.
“Hasn’t anyone seen Terminator, for Christ’s sake?” says Allen, 53, whose work over the past two decades has been played on major radio stations like Classic FM and Heart FM.
“I think it could very easily take over.”
AI started by depleting requests for voiceovers in company training videos, but Allen recently lost a potential radio client who instead bought the first AI advert he’s ever heard that’s good enough for broadcast.
“It was scarily good,” says Allen, who lives near Birmingham. “No one would know.”
The cost to the client? £11.99. Voice actors would expect £1,000.
“There’s no way anybody can compete.”
Image: Pic: iStock
Shifting sands forming another job desert?
Not according to Oxford’s Labour economist Fabian Stephany, who was keen to “challenge the dystopian narrative”.
“It is very rare for a new technology to completely replace an entire profession,” says Stephany.
The only exceptions are jobs defined by a single task without any complexity, like bowling alley pinsetters or the translators at the top of Microsoft’s table, he says.
There’s complexity in Allen’s job, like creating video and TV soundtracks and mixing audio, but he’s still nervous.
“The AI subscription can mix for you too, so that’s production houses everywhere – we’re no longer needed. That’s quite scary.”
He adds: “I won’t be doing this in 10 years’ time.”
Microsoft researcher Kiran Tomlinson says AI “may prove to be a useful tool for many occupations” and “the right balance lies in finding how to use the technology in a way that leverages its abilities while complementing human strengths and accounting for people’s preferences”.
In January, Sir Keir Starmer said there was “barely an aspect of our society that will remain untouched” by AI in the coming years.
The technology is mentioned at least 126 times in the government’s industrial strategy for the tech sector, focusing heavily on its potential benefits.
Insufficient attention is being paid to its disruption, says Zhu. Why is Microsoft publishing reports on job exposure, but not the government? Where is the guidance on how employers and employees should adapt?
“The government should play some important role here, and they’re not,” she says.
Recalling how laid-off steelworkers were left to fend for themselves in the 20th century, Stephany warns it is “crucial to not make the mistakes of the past again”.
Allen couldn’t agree more: “All jobs under threat of AI need to be protected. Because otherwise, how the hell do people earn money?”
The government says it is putting people “at the heart” of its AI plans.
“That includes partnerships with leading tech firms to help us deliver AI skills training to 7.5 million workers, and initiatives to bring digital skills and AI learning into classrooms and communities,” a spokesperson says.
“This will provide training to people of all ages and backgrounds and is backed by £187m.”
They say “thousands of jobs” will be created by AI Growth Zones, areas earmarked for AI data centres where the state will support big tech companies with access to power and planning.
Image: Keir Starmer announces the TechFirst programme teaching school pupils AI in June. Pic: PA
What can you do for yourself?
Workers should be concerned if they’re not trying to use AI, says the consultant.
CVs with AI skills have so far been consistently favoured by a group of 2,000 recruiters observed by Fabian Stephany in an ongoing study.
“If a worker is willing to invest in their skill set, in developing their profile, they should not be worried at all,” they say.
Almost half (45%) of global employers consider AI competency to be a core skill, according to the World Economic Forum.
LinkedIn data shows AI-related skills on member profiles rose 65% year-on-year in 2024.
Job postings on Indeed.com containing AI terms have risen by 170% since the end of June 2023 – albeit from a low starting point (1.7% to 4.6% by 31 August).
“If you’re willing to learn skills that allow you to integrate AI into the job that you’re currently doing, you will probably not only be doing fine, but you might actually have a big career boost ahead of you,” Stephany adds.
In a separate study of 10 million job vacancies in the UK, he found jobs asking for AI skills paid 23% more – a salary boost greater than that expected from a master’s degree (20%).
The best starting point is creating a free account with AI chatbots like ChatGPT, Claude or Gemini, says the AI consultant.
“Log into one of them, provide it a pretty detailed description of who you are, what you do day-to-day, both in your job and potentially in your personal life, and ask it how it can help.
“Right now, that can mean that you do your job better, which gets you promoted.”
Or maybe not.
In the past few months, writer Joe Turner has seen some clients make a sheepish return.
“I see an influx of new jobs coming in and people are now requesting no AI content at all,” he says.
Clients have found its hollow tropes and generic mannerisms carry the unmistakable mark of a “soulless machine”.
“It’s called AI, but it’s not artificial intelligence. It’s just a database of words with reasoning models,” he concludes.
“It puts the words in the right order, but at the end of the day, it means nothing.”