The Vogtle nuclear power plant is located in Burke County, near Waynesboro, Georgia in USA. Each of the two existing units have a Westinghouse pressurized water reactor (PWR), with a General Electric turbine and electric generator, producing approximately 2,400 MW of electricity. Two Westinghouse made AP 1000 reactors are under construction here.
Pallava Bagla | Corbis News | Getty Images
Venture capitalists in Silicon Valley and other tech hubs are investing money in nuclear energy for the first time in history. That’s changing its trajectory and pace of innovation.
“There’s not been a resurgence of nuclear power, ever, since its heyday in the late 1970s,” Ray Rothrock, a longtime venture capitalist who has personal investments in 10 nuclear startups, told CNBC.
Now, that’s changing. “I have never seen this kind of investment before. Ever.”
Jacob DeWitte, CEO of micro-reactor startup Oklo, says the landscape has changed dramatically since he started raising money in 2014, when he was a part of the Y Combinator startup incubator.
“More investors are interested, more investors are excited by the space, and they’re getting smarter to do the diligence and know what to do here — which is good,” DeWitte told CNBC.
This surge of private investment will be a positive for the industry, agrees John Parsons, an economist and lecturer at MIT.
“I think having fresh perspectives is really good,” Parsons told CNBC. Nuclear energy is “a very complex science, and it’s been supported by the federal government and at these national labs. And so that’s a very small circle of people. And when you broaden that circle, you get a lot of new minds, different thinking, a variety of experiments.”
In any industry, there can be a “groupthink” or “narrowness” in the way things are done over time, Parsons said. With private investment in the space, “there will be out-of-the-box thinking,” he said. “Maybe that out-of-the-box thinking doesn’t produce anything useful. Maybe it turns out that the old designs are the best. But I think it’s really wonderful to have the variety of takes.”
Not everyone is so optimistic that the recent influx of venture dollars will lead to progress.
“Investors have often invested in stupid things that didn’t work,” Naomi Oreskes, a professor of the history of science at Harvard University, told CNBC. “Because the reality is that in a 75-year history of this technology, it has never been profitable in a market-based system.” If investors are putting money into nuclear now, that’s because they think they can make money, and “I can only think they believe they will make money because they think that there’s a big opportunity to have the federal government pick up a big part of the tab,” Oreskes said.
Pitchbook’s private investment data for nuclear technology data includes both fusion and fission.
Chart courtesy Pitchbook.
Nuclear investment by the numbers
From 2015 to 2021, total venture capital deal flow in the United States increased 54% in terms of deals closedand 294% by dollar value, according to data compiled by private capital market research firm Pitchbook for CNBC. In that same time, climate investing deal flow in the United States jumped by 214% in terms of volume and 1,348% by dollar value.
In the nuclear space, investment rose even faster — 325% by volume and 3,642% by dollar value, according to Pitchbook.
Some of the rapid pace of increase in investment in the nuclear sector is explained by its starting point — virtually zero.
The venture market slowed overall in 2022, and nuclear investment is no exception. Concerns about the war in Ukraine, inflation, a wave of layoffs and murmurs of a recession have made investors nervous in the public markets and private alike.
Pitchbook includes companies developing technologies to mitigate or adapt to climate change in this category. Examples include renewable energy generation, long duration energy storage, the electrification of transportation, agricultural innovations, industrial process improvements, and mining technologies.
Chart courtesy Pitchbook
“At the beginning of the year, we were looking at a much different financial paradigm for nuclear startups seeking funding. Now, following a war, and inflationary related forces, the fundraising market is just not what it was earlier and that is challenging for everyone seeking funding and support, nuclear or otherwise,” Brett Rampal, a nuclear energy expert who evaluates investment opportunities and consults for nuclear startups, told CNBC.
More than $300 billion poured into the venture capital industry in 2021. Rothrock expects to see more like $160 billion in 2022.
“I’m sure that some funds that pull back may never come back,” Rothrock said. But most investors who are putting money into a nuclear company understands that it will not be a quick investment, Rothrock told CNBC. “Entrepreneurs and investors at the level we are talking for nuclear are playing the long game, they have to. These projects will take time to mature and to generate real cash flows.”
Also, the Inflation Reduction Act that President Joe Biden signed into law in August, which includes $369 billion in funding to help combat climate change, has given nuclear investors a very significant positive signal, Rampal told CNBC.
“The IRA investment and production tax credits are not nuclear specific credits, they’re clean energy credits that nuclear is now considered a part of, and that sends a real important message to people and investors that would consider this space,” Rampal said. Similarly important, the European Union voted in July to keep some specific uses of nuclear energy (and natural gas) in its taxonomy of sustainable sources of energy in some circumstances, according to Rampal.
Total venture capital deal activity, according to Pitchbook data, for the last five years.
Chart courtesy Pitchbook.
The VC approach to nuclear
The nuclear power industry in the United States launched as a government project after the U.S. built the first atomic bombs during World War II. In 1951, a nuclear reactor produced electricity for the first time in Idaho at the National Reactor Testing Station, which would become the Idaho National Laboratory.
In the 1960s and 1970s, large conglomerates constructed big nuclear power plants, and those projects often ran over budget. “As a consequence, most of the utilities that undertook nuclear projects suffered ratings downgrades—sometimes several downgrades—during the construction phase,” according to a 2011 report from the Congressional Budget Office. Also, the Three Mile Island accident in 1979 raised public fears about safety and put a damper on construction.
However, in recent years, private investors and venture capitalists have been putting money into nuclear startups, driven by a newfound sense of urgency to respond to climate change, as nuclear energy releases no greenhouse gases. There’s also the allure of funding underdog companies with huge upside.
The venture capital model is based on big bets — venture capitalists spread their money across many companies. Most are expected to fail or maybe break even, but if one or two companies get enormous, they more than cover the cost of all those losses. This is the investing model that built Silicon Valley stalwarts like Apple, Google and Tesla.
Some venture capitalists are especially excited about fusion. It’s the type of nuclear energy that powers stars, and it generates no long-lasting radioactive waste — but so far, it’s proven fiendishly difficult to create a lasting fusion reaction on Earth and impossible to generate enough energy for commercial generation.
“It’s far better than nuclear fission,” investor Vinod Khosla told CNBC in October. “It’s far better than coal and fossil fuels for sure. But it’s not ready. And we need to get it ready and build it.”
Khosla isn’t the only one. The private fusion industry has seen almost $5 billion in investment, according to the Fusion Industry Association, and more than half of that has been since since the second quarter of 2021, Andrew Holland,CEO of the association, told CNBC.
Installation of one of the giant 300-tonne magnets that will be used to confine the fusion reaction during the construction of the International Thermonuclear Experimental Reactor (ITER) on the Cadarache site on September 15, 2021.
Jean-marie Hosatte | Gamma-rapho | Getty Images
Others are excited about new advances in nuclear fission, the more traditional type of nuclear power based on breaking atomic nuclei apart, like DCVC founder Zachary Bogue, who invested in micro-nuclear reactor company Oklo.
“Advanced nuclear fission is a quintessential deep-tech venture capital problem,” Bogue told CNBC in September. There is technical and regulatory risk, but if those problems are solved, “there are just massive-scale returns … all of those elements are a perfect recipe for venture capital.”
While these bets seem expensive and risky compared with venture capital’s recent focus on software and consumer tech, they’ll still bring a faster and more agile approach than the old-line nuclear industry.
Take micro-reactors.
“These are going to be very expensive at first. But the goal is to find something that is a product that’s much more flexible, can go on to the grid in many more different places and serve different functions, and go off grid also,” explained MIT’s Parsons.
Similarly, fusion startups say they will generate energy much faster than government research projects like ITER, which has already been in progress since 2007.
This quick-turn approach to investment is spurring experimentation. New generations of nuclear reactors will have different sizes, different coolants and different fuels, explained Matt Crozat, senior director of policy development at the Nuclear Energy Institute. Some reactors are being designed for companies or communities in isolated areas, for example. Others are being made to operate at high temperatures for industrial processes, Crozat told CNBC.
“It really is expanding the range of what nuclear can mean,” Crozat said. Many won’t succeed, but time and the market will figure out what’s needed and what’s possible, he said.
Because venture investors are hungry for returns, this also spurs nuclear startups to chase interim revenue streams as they’re getting their big-bet technology up and running.
But critics say venture capitalists are ignoring the troubled history of nuclear power as a business.
“Investors have forgotten or are ignoring the lessons from earlier generations of nuclear plants which cost 2 to 3 times as much to build and took years longer than was promised by the vendors,” Schlissel told CNBC. For instance, a project to put two new reactors on the Vogtle power plant in Georgia was originally estimated to be $14 billion and ended up costing more than $34 billion and taking six years longer to complete than expected, he said.
15 November 2022, Egypt, Scharm El Scheich: A nuclear symbol is displayed at a pavilion of the International Atomic Energy Agency IAEA at the UN Climate Summit COP27. Photo: Christophe Gateau/dpa
Harvard’s Oreskes says the nuclear industry is a “technology with a long history of broken promises,” and she is skeptical of the sudden investor interest.
“If you were my daughter, and you had a boyfriend that had made repeated promises to you over months, years, decades, constantly breaking them, I would say, ‘Do you really want to be with this guy?'”
She’s not categorically anti-nuclear, and supports the continued operation of nuclear power plants that already exist. But she’s particularly skeptical of fusion, which has been promised to be “just around the corner” for decades, and says this new round of investments in fusion “doesn’t pass the laugh test.”
Ultimately, the new crop of nuclear startups has to figure out how to create nuclear energy in a cost-competitive way, or nothing else matters, says Rothrock.
“More money means more startups and to me that means more shots on goal (improving odds of success),” he told CNBC.
“The issue in nuclear is economics. Plants are complicated and take a while to build. Some of these new startups are tackling those issues making them more simple and thus cheaper. No one will buy an expensive power plant, especially a nuclear plant. Economics drives it all.”
Tesla has been forced to reimburse a customer’s Full Self-Driving package after an arbitrator determined that the automaker failed to deliver it.
Tesla has been promising its car owners that every vehicle it has built since 2016 has all the hardware capable of unsupervised self-driving.
The automaker has been selling a “Full Self-Driving” (FSD) package that is supposed to deliver this unsupervised self-driving capability through over-the-air software updates.
Almost a decade later, Tesla has yet to deliver on its promise, and its claim that the cars’ hardware is capable of self-driving has been proven wrong. Tesla had to update all cars with HW2 and 2.5 computers to HW3 computers.
Tesla is now attempting to deliver its promise of unsupervised self-driving on HW4 cars, which have been in production since 2023-2024, depending on the model. However, there are still significant doubts about this being possible, as the best available data indicate that Tesla only achieves about 500 miles between critical disengagements with the latest software on the hardware.
On the other hand, many customers are losing faith in Tesla’s ability to deliver on its promise and manage this computer retrofit situation. Some of them have been seeking to be reimbursed for their purchase of the Full Self-Driving package, which Tesla sold from $8,000 to $15,000.
A Tesla owner in Washington managed to get the automaker to reimburse the FSD package, but it wasn’t easy.
The 2021 Model Y was Marc Dobin and his wife’s third Tesla. Due to his wife’s declining mobility, Dobin was intrigued about the FSD package as a potential way to give her more independence. He wrote in a blog post:
But FSD was more than hype for us. The promise of a car that could drive my wife around gave us hope that she’d maintain independence as her motor skills declined. We paid an extra $10,000 for FSD.
Tesla’s FSD quickly disillusioned Dobin. First, he couldn’t even enable it due to Tesla restricting the Beta access through a “safety score” system, something he pointed out was never mentioned in the contract.
Furthermore, the feature required the supervision of a driver at all times, which was not what Tesla sold to customers.
Tesla doesn’t make it easy for customers in the US to seek a refund or to sue Tesla as it forces buyers to go through arbitration through its sales contract.
That didn’t deter Dobin, who happens to be a lawyer with years of experience in arbitration. It took almost a year, but Tesla and Dobin eventually found themselves in arbitration, and it didn’t go well for the automaker:
Almost a year after filing, the evidentiary hearing was held via Zoom. Tesla produced one witness: a Field Technical Specialist who admitted he hadn’t checked what equipment shipped with our car, hadn’t reviewed our driving logs, and didn’t know details about the FSD system installed on our car, if any. He hadn’t spoken to any sales rep we dealt with or reviewed the contract’s integration clause.
There were both a Tesla lawyer and an outside counsel representing Tesla at the hearing, but the witness was not equipped to answer questions.
Dobin wrote:
He was a service technician, not a lawyer or salesperson. But that’s who Tesla brought to the hearing. At the end, I genuinely felt bad for him because Tesla set him up to be a human punching bag—someone unprepared to answer key questions, forced to defend a system he clearly didn’t understand. While I was examining him, a Tesla in-house lawyer sat silently, while the company’s outside counsel tried to soften the blows of the witness’ testimony.
He focused on Tesla’s lack of disclosure regarding the safety score and the fact that the system does not meet the promises made to customers.
The arbitrator sided with Dobin and wrote:
The evidence is persuasive that the feature was not functional, operational, or otherwise available.”
Tesla was forced to reimburse the FSD package $10,000 plus taxes, and pay for the almost $8,000 in arbitration fees.
Since Tesla forces arbitration through its contracts, it is required to cover the cost.
Electrek’s Take
This is interesting. Tesla assigned two lawyers to this case in an attempt to avoid reimbursing $10,000, knowing it would have to cover the expensive arbitration fees – most likely losing tens of thousands of dollars in the process.
It makes no sense to me. Tesla should have a standing offer to reimburse FSD for anyone who requests it until it can actually deliver on its promise of unsupervised self-driving.
That’s the right thing to do, and the fact that Tesla would waste money trying to fight customers requesting a refund is really telling.
Tesla is simply not ready to do the right thing here, and it doesn’t bode well for the computer retrofits and all the other liabilities around Tesla FSD.
FTC: We use income earning auto affiliate links.More.
After hitting a major milestone on Monday, BYD claimed it’s about to unleash “the largest-scale smart driving OTA in history.”
BYD preps for the largest-scale software update
BYD announced on Weibo that there are now over 1 million vehicles on the road with its God’s Eye smart driving system.
The milestone comes after it upgraded 21 of its top-selling vehicles with the smart driving tech in February, at no extra cost. Even its most affordable EV, the Seagull, which starts at under $10,000 (69,800 yuan), got the upgrade.
BYD didn’t reveal any specifics, only promising “it is safer and smarter.” The Chinese EV giant has three different “God’s Eye” levels: A, B, and C.
Advertisement – scroll for more content
The highest, God’s Eye A, is typically reserved for BYD’s ultra-luxury Yangwang brand, which utilizes its DiPilot 600 smart cockpit with three LiDARs.
God’s Eye B is used for other luxury and higher-end models, including those under Denza, which utilize DiPilot 300 and one or two LiDARs.
The base God’s Eye C system, used for BYD brand models, includes 12 cameras, five wave radars, and 12 ultrasonic radars, all supported by DiPilot 100.
Last week, BYD’s luxury off-road brand, Fang Cheng Bao, launched a limited-time offer for Huawei’s Qiankun Intelligent Driving High-end Function Package. The discount cuts the price from 32,000 yuan ($4,500) to just 12,000 yuan ($1,700).
BYD Seagull EV testing with God’s Eye C smart driving system (Source: BYD)
After selling another 382,585 vehicles in June, BYD now has over 2.1 million in cumulative sales in the first half of 2025, up 33% from last year.
With the “largest-scale smart driving” update coming soon, BYD’s vehicles are about to gain new functions and safety features. Check back soon for more details.
BYD claims it’s “capable of leading the transformation and popularization of intelligent driving” with over 5,000 engineers dedicated to the field. As the world’s largest NEV maker, BYD said it’s committed to transforming the auto industry with safer and more sustainable solutions for global markets.
FTC: We use income earning auto affiliate links.More.
Kia’s electric SUV is a hit in the UK. The EV3 was the most popular retail EV through the first half of 2025, pushing Kia to become the UK’s third top-selling car brand so far this year.
Kia EV3 leads as the UK’s most popular retail EV
The EV3 is Kia’s fastest-selling EV in the UK and a massive part of the brand’s success this year. Kia said the compact electric SUV contributed to its best-ever June, Q2, and first half EV registrations so far this year.
In January, the EV3 “started with a bang,” racing out to become the UK’s most popular retail EV. The EV3 was the best-selling retail EV in the UK and the fourth best-selling EV overall in the first quarter, including commercial vehicles.
Through the first half of the year, the Kia EV3 maintained its crown as the UK’s most popular EV with 6,293 registrations.
Advertisement – scroll for more content
The EV3 starts at £33,005 ($42,500) as the ‘brand’s most affordable EV yet.” It’s available with two battery packs: 58.3 kWh or 81.48 kWh, providing a WLTP range of up to 430 km (270 miles) and 599 km (375 miles), respectively.
Kia EV3 (Source: Kia)
Kia sold 31,643 electrified vehicles in the first half of 2025. Although this includes fully electric vehicles (EVs), plug-in hybrids (PHEVs), and hybrids (HEVs), it still accounts for over half of Kia’s total of 62,005 registrations.
Kia EV3 (Source: Kia)
After opening orders for the EV4 last week, Kia’s first electric hatchback, the brand expects to see even more demand throughout 2025. With up to 388 miles WLTP range, it’s also the longest-range Kia EV to date.
Next year, Kia will introduce the entry-level EV2, which will sit below the EV3 in Kia’s lineup. Kia is looking to add an even more affordable EV to sit below the EV2. It will start at under $30,000 (€25,000), but we likely won’t see it until closer toward the end of the decade.
FTC: We use income earning auto affiliate links.More.