Demis Hassabis is one of the defining figures of the AI era. Google recognized his potential early, buying DeepMind just four years after he founded it.

What once began as a small London-based company now sits at the heart of Google’s fight to lead in artificial intelligence, as the company races to compete with a growing field of global rivals including OpenAI and Anthropic.

But a leadership shakeup announced on Wednesday, which saw Hassabis step back from being CEO to become chair of Google DeepMind and parent company Alphabet‘s chief scientist leave, concerned investors and the share price dropped following the announcement.

I have covered DeepMind for more than a decade and recently sat down with Hassabis for an exclusive interview. What may have come as a surprise to many, to me, feels like a natural progression.

But the move also exposes the tension at the center of Google’s AI strategy: Hassabis wants to use AI to transform science, while Google needs to commercialize, monetize and scale the technology fast enough to justify the billions it is spending.

Hassabis has been very vocal about his goals with AI. In our interview on CNBC’s podcast “The Tech Download” that was recorded in December, Hassabis said he spent his whole career on AI because it will eventually become “the ultimate tool for science.”

Over the years, under Hassabis’ leadership, DeepMind has made some significant breakthroughs. One of those achievements was AlphaGo, a computer system that took on the best player in the world at the ancient Asian board game Go and won. Another is AlphaFold, an AI software that can predict protein structure folding, which is key for understanding diseases and drug discovery.

Hassabis told me none of this would have been possible without Google’s backing and the “amount of compute that they could bring to play,” referring to the tech giant’s ample computing power driven by its own chips.

The DeepMind co-founder has been instrumental in pushing the science from within DeepMind at Google. He has also pursued artificial general intelligence, or AGI — AI that can match human capabilities across most cognitive tasks. Hassabis on Wednesday said he believes AGI is “close at hand” in a blog post on Wednesday accompanying the news of his new role.

Hassabis said his motivation for handing over day-to-day operations is to work on strategic and global AGI matters, advise his successor, the new senior vice president of Google DeepMind, Koray Kavukcuoglu, and dedicate time to Isomorphic Labs, a company he co-founded that is dedicated to AI drug discovery.

The move from the company’s co-founder “isn’t out of character,” said Sebastian Mallaby, who wrote “The Infinity Machine,” about Hassabis and DeepMind.

“Demis cared about safety enough that he sold DeepMind to Google, not to Facebook, even though Facebook offered more money. He cared enough about safety that he fought a three-year battle with Alphabet to get external oversight…

Google’s AI divisions are getting reshuffled, the search giant announced on Wednesday, with chief scientist Jeff Dean leaving the company after 27 years.

Demis Hassabis, the CEO of Google DeepMind, is moving into a chairman role of that unit and also assuming the title chief scientist of parent company Alphabet, according to a memo from CEO Sundar Pichai that was posted to Google’s blog.

Alpahbet shares fell about 4% after the announcement.

Dean, a pioneer in artificial intelligence who’s been credited with some of Google’s most important technical breakthroughs, is starting his own company along with Google senior fellow Sanjay Ghemawat, the post said. The departure is on friendly terms and Google will invest in his startup, a representative said.

“After an incredible 27-year run, Jeff Dean is at a moment where he wants to try something new, and we’re excited to support him in that,” Pichai said in the post. He added that Dean and Ghemawat will be working to “accelerate discoveries” in machine learning, science and engineering.

The shakeup, which includes promoting DeepMind technology chief Koray Kavukcuoglu to head of the AI division, comes as Google navigates a rapidly evolving AI industry, with the company trying to compete against OpenAI and Anthropic in developing the most cutting-edge frontier models while also pouring money into infrastructure so that its cloud division can serve customers as well as its own workloads.

In the latest quarter, Google turned cash flow negative for the first time on record due to its capital expenditures, while forecasting full-year capex of up to $205 billion. Kavukcuoglu will lead the development of Gemini 4, the company’s next major AI model, and will report to Pichai.

Hassabis co-founded DeepMind and joined Google when the search giant purchased the lab in 2014. In recent years, he’s been leading nearly all of Alphabet’s foundational AI work.

“I’ve decided that now is the right time for me to hand over my day-to-day operational responsibilities at GDM, so that I have the time and space to focus on the big picture and help influence what is to come to the best of my ability,” Hassabis said in a note to employees.

He added he would work with Pichai on “strategic and global” matters related to AI.

In a followup post on X, Dean said his new startup will be called Discovery Loop, and that it will be organized as a public benefit corporation focused on AI for science and engineering.

From a growth perspective, Google has been doing just fine on AI. The company’s cloud division, run by Thomas Kurian, is expanding at a much faster clip than larger rivals Amazon Web Services and Microsoft Azure.

Cloud revenue at Google soared 82% in the second quarter to $24.8 billion. AWS reported revenue growth of 37% in the quarter, while Azure sales climbed 43%.

Pichai said on the earnings call that cloud expansion was driven by AI infrastructure and AI solutions, noting and was helped by demand for the company’s homegrown…

Anthropic CEO Dario Amodei and Google DeepMind CEO Demis Hassabis called for a U.S.-led coalition to shape rules and standards around artificial intelligence at a meeting with tech leaders and heads of state, including President Donald Trump, CNBC has learned.

The closed-door lunch meeting took place on Wednesday at the G7 summit in Évian-les-Bains, France. Amodei and Hassabis both proposed international cooperation on AI, with the U.S. taking the lead, to protect against risks associated with the emerging technology, according to two people with knowledge of the matter who asked not to be named because they weren’t authorized to discuss the meeting.

Canadian Prime Minister Mark Carney agreed that the U.S. could lead an AI coalition, according to one of the people and another person familiar with the discussions.

Anthropic declined to comment on the G7 meeting. Google DeepMind and the Office of the Prime Minister of Canada didn’t immediately respond to requests for comment.

The gathering follows the release of increasingly powerful AI models with cyber capabilities so advanced that some industry experts have raised concerns that they can cause major disasters if left in the wrong hands. Most recently, Anthropic disabled access to its newest models, Fable 5 and Mythos 5, on Friday after the U.S. government imposed export controls on the models, citing national security concerns.

In addition to Amodei and Hassabis, OpenAI CEO Sam Altman attended Wednesday’s meeting, which included about a dozen tech executives. Leaders of G7 countries were also involved.

Alongside President Trump, Secretary of the Treasury Scott Bessent, Secretary of Commerce Howard Lutnik and Secretary of State Marco Rubio represented the U.S.

Amodei said in his address to the group that areas of international cooperation should include structured access to frontier models, and trade of chips and critical components that excludes China, one person said. Amodei also said that countries should cooperate to address the risk of AI in cyber, bioterrorism and areas of intelligence, the person added.

Anthropic remains locked in negotiations with the Trump administration after the export controls on its latest models went into effect late Friday.

Altman, in his comments, called for “an international forum for discussion that establishes globally accepted standards for testing, provides expert and impartial analysis of capabilities and risks, and serves as a venue for cooperation among nations,” according to a briefing from OpenAI.

Last month, OpenAI announced that GPT-5.5 Cyber, a variation of its latest model, was rolling out in a limited preview capacity to vetted cybersecurity teams.

OpenA’s global affairs chief Chris Lehane, who attended…

Chiefs of the world’s leading AI companies are descending on the G7 conference in France Wednesday, in a sign of their growing geopolitical influence as artificial intelligence rises to the top of the global agenda.

CEOs including OpenAI’s Sam Altman, Anthropic’s Dario Amodei, Google DeepMind’s Demis Hassabis, alongside around a dozen other tech leaders, will take part in a lunch meeting at the summit in Evian on Wednesday.

Frontier AI risks, infrastructure and sovereignty are all expected to be discussed at the conference. The protection of children online will also be a key part of the discussions, The Élysée Palace, the official residence of the president of France in Paris, said in a press briefing on Thursday.

Other tech chiefs including France-based Mistral’s Arthur Mensch, Canada’s Cohere CEO Aidan Gomez, Italian company Domyn’s Uljan Sharka, U.K. AI scaleup Synthesia’s Victor Riparbelli and German-based Black Forest Labs’ Robin Rombach will also be present at the lunch. Salesforce‘s Marc Benioff, Meta‘s Alex Wang, alongside the founders of Indian AI company Sarvam and Japan’s Sakana are also pegged to attend.

“It just shows that in order to make credible commitments on AI, heads of state now need the cooperation, if not endorsement, of a handful of private sector executives actually building the technology,” Jessica Brandt, senior fellow for technology and national security at the Council on Foreign Relations (CFR), told CNBC.

“We’re seeing a shift in who gets a seat at the table and a signal of where power sits.”

The G7 summit — which features the U.S., U.K., Canada, France, Germany, Italy, Japan and the EU — comes as Anthropic remains locked in negotiations with the U.S. administration after Washington imposed export controls on the AI lab’s Fable 5 and Mythos 5 models amid national security concerns.

Recent announcements of powerful AI models with advanced cyber capabilities, including Anthropic’s Mythos and OpenAI’s GPT-5.5 Cyber, have brought a wave of concerns from businesses and governments around digital security weaknesses.

The release of Mythos marked an “inflection point” in AI development, Cameron Kerry, a visiting fellow at the Brookings Institution, told CNBC, adding that it led the Trump administration to consider regulating the technology.

U.S. export controls on Anthropic’s models have “changed everything,” said Emerson Brooking, senior fellow at the Atlantic Council.

“Multiple G7 nations have previously alluded to the need for sovereign AI investment, but there was always an assumption that this would take place alongside access to the U.S. tech stack,” he told CNBC. “Now the U.S. has indicated a willingness to cut off the G7 and even treaty allies from certain AI capabilities.”

For tech bosses, a seat at the table during the G7 represents a key opportunity to influence policy debates at the highest level.

“It seems the firms expect to come away with a package of voluntary commitments — youth…

The latest JPMorgan Summer Reading List, an annual favorite among the wealthy, includes books on artificial intelligence, longevity, Keith Haring and lemon recipes.

The list is compiled each year with suggestions from JPMorgan client advisors around the world and has become a popular guide for the wealthy looking for good summer beach reads. This year’s list includes 14 titles, reflecting the top issues and concerns of its clients.

AI tops the list, with two books on the topic. Sebastian Mallaby’s “The Infinity Machine” profiles Google DeepMind CEO and co-founder Demis Hassabis and the battle for AI breakthroughs. Josh Tyrangiel’s “AI for Good” looks at how the technology is already helping to solve everyday challenges.

The list also features books on leadership and performance. Cognitive scientist George Newman’s book, “How Great Ideas Happen,” is about how to develop skills for discovery. Neuroscientist and physician Dr. Tommy Wood’s “The Stimulated Mind,” explains strategies on how to keep the brain adaptable and focused.

With sports investing growing among the wealthy, the list also includes NBA analyst Ric Bucher’s book, “Coachable,” which gleans lessons from conversations with former NBA star Michael Jordan, ex-NFL quarterback Tom Brady, former WNBA great Diana Taurasi and others.

Darin Oduyoye, the chief communications officer of J.P. Morgan Asset and Wealth Management, who oversees the list, said family offices are especially interested in books on generational transition, since many are managing the transfer of wealth to their children or family. Of particular interest to family offices is the book “Mattering,” by Jennifer Breheny Wallace, Oduyoye said.

“Her message is resonating with next-generation family leaders — centered on purpose, identity, and how rising decision-makers define what they want to pursue — especially as families navigate leadership transition and legacy,” he said.

Here is the full list of books:

Alphabet shares started 2025 with investors questioning whether Google could keep up with ChatGPT maker OpenAI in the AI race. By year’s end, the stock had notched its best performance since 2009.

Google got its AI mojo back. Much of that was driven out of DeepMind, the British company Google acquired in 2014 for around £400 million.

In a wide-ranging interview for CNBC’s new podcast, The Tech Download, DeepMind’s founder and CEO Demis Hassabis called it “the engine room” of Google’s AI efforts, adding that changes had been made to enable the tech giant to rapidly roll out AI products amid a “ferocious competitive environment.”

Hassabis said he talks to Google CEO Sundar Pichai “every day,” underscoring how close the two executives are working to innovate quickly.

“All the AI technologies is done by this group … and then it’s diffused across all of these incredible products right across Google,” Hassabis told The Tech Download, which launched on Friday.

“And the last couple of years, we’ve been building that backbone, so not just the models, but also … architecting the entire infrastructure of Google so that … these things can ship incredibly quickly.”

This could be key for Google as it faces another year of competition from OpenAI as well as a plethora of other players from Amazon to Perplexity and Anthropic.

“It’s a ferocious competitive environment at the moment,” Hassabis said. He added “many” veterans who’d been in tech for “20, 30 years,” had told him this was “the most intense environment they’ve ever seen, perhaps ever in the technology industry.”

In 2023, Google made a key change to combine its Google Brain research division with DeepMind, a move that laid the foundation for its success with the company’s flagship AI assistant Gemini. Other key shifts, such as promoting executive Josh Woodward to run Gemini, played their part.

When OpenAI launched ChatGPT in November 2022, Google was playing catch-up. Product missteps with its AI tools along the way, particularly in 2024, reinforced the industry’s impression that Google was struggling to compete.

Hassabis said the company’s issue wasn’t inventing tech. Transformers, a key architecture that underpins large language models, were created by Google researchers after all. The company’s issue was “maybe” that it was “a little bit slow to commercialize it and scale it,” Hassabis continued.

“That’s what OpenAI and others did very well,” he added.

“The last two, three years, I think we’ve had to come back to almost our startup or entrepreneurial roots and be scrappier, be faster, ship things really quickly and sort of make really rapid progress,” Hassabis said.

The DeepMind CEO said the…

China’s artificial intelligence models may be just “a matter of months” behind U.S. and Western capabilities, Demis Hassabis, the CEO of Google DeepMind told CNBC.

The assessment from the head of one of the world’s leading AI labs and a key driver behind Google’s Gemini assistant, runs counter to views that have suggested China remains far behind.

Speaking on CNBC’s new podcast, The Tech Download, which launched on Friday, Hassabis said Chinese AI models are closer to U.S. and Western capabilities “than maybe we thought one or two years ago.”

“Maybe they’re only a matter of months behind at this point,” Hassabis told The Tech Download.

About a year ago, Chinese AI lab DeepSeek came out with a model that sent shockwaves through markets because of its strong performance that was built on less-advanced chips and at a lower cost than American alternatives.

While DeepSeek has released new models since, and the shock factor has worn off, China’s tech giants like Alibaba and startups such as Moonshot AI and Zhipu have also released very capable models.

Still, Hassabis said that while China could play catch up, the country’s companies are yet to prove their ability to create AI breakthroughs.

“The question is, can they innovate something new beyond the frontier? So I think they’ve shown they can catch up … and be very close to the frontier … But can they actually innovate something new, like a new transformer … that gets beyond the frontier? I don’t think that’s been shown yet,” Hassabis said.

The transformer was a scientific breakthrough made by Google researchers in 2017 that underpins the large language models that have been developed by AI labs in recent years, including those powering products like OpenAI’s ChatGPT and Google’s Gemini.

Other top technology figures have also given credit to China’s progress. Nvidia CEO Jensen Huang said last year that the U.S. is “not far ahead” in the AI race.

“China is well ahead of us on energy. We are way ahead on chips. They’re right there on infrastructure. They’re right there on AI models,” Huang said.

China’s technology firms face a number of challenges, with access to critical technology among the biggest. There is a U.S. export ban in force on leading-edge semiconductors from Nvidia that are required to train more advanced AI models.

The White House has indicated that it would approve sales of Nvidia’s H200 chip to China, a more advanced semiconductor than the country recently had access to. However, it is not Nvidia’s top-of-the-range product.

Homegrown chip firms like Huawei have looked to fill the gap, but their performance still lags behind Nvidia’s offering.

Some analysts have suggested that over the longer term, the lack of access to Nvidia chips in China could mean the gap between U.S. and…