
Inside Europe’s tech hubs: Does London have what it takes to remain No. 1?
More Videos
Published
11 months agoon
By
adminThe U.K. is the No. 1 country in Europe for tech by many accounts.
In 2023, British startups raised $21.3 billion, their third-highest total on record, according to a report from Dealroom and HSBC Innovation Banking.
It’s the No. 1 location in Europe for funding — with France coming in second, raising $9.2 billion last year — although it remains well behind the U.S. and China.
In the mid-2010s, the U.K., and London in particular, saw a boom of startups in financial technology, or fintech, given the city’s importance to the global financial services market. Since then, major companies across different sectors have developed, spanning food delivery to cybersecurity.
The U.K. was also the birthplace of two, now foreign-owned, major tech firms: chip designer Arm and artificial intelligence firm DeepMind.
This startup ecosystem, along with its top-tier universities and international investors, enabled the U.K. to become a major tech hub.
But its status could be under threat.
The country is facing a slew of challenges, including the fallout from its EU exit, which formally took place in 2020, and very low numbers of tech stock market listings.
Meanwhile, other European countries are making a play to become the No. 1 tech destination, namely France.
In the latest episode of CNBC Tech’s “Beyond the Valley” podcast — which you can listen to above — Tom Chitty and I discuss whether the U.K. can maintain its lead as Europe’s top technology hub and the challenges it faces ahead.
If you have any thoughts on this or previous episodes, please email us at beyondthevalley@cnbc.com.
You can subscribe to “Beyond the Valley” by clicking the links below to your chosen platform:
Here is a transcript of the “Beyond the Valley” episode released on June 11, 2024. It has been edited for clarity and brevity.
Tom Chitty
Since the start of the year, Arjun has traveled around Europe speaking to the leaders and innovators in tech. And now we’re going to dig a little deeper. Over the next few weeks, we’ll find out what’s happening in some of Europe’s key tech hubs with the help of a special guest, who will help us understand its inner workings and the challenges it may face. This week, we start with the U.K. Arjun, before we introduce our guest, it was your suggestion to do this series. What do you hope our listeners will learn from it?
Arjun Kharpal
Are you saying if it goes badly, it’s my suggestion?
Tom Chitty
You’re to blame.
Arjun Kharpal
I take that. I’m hoping that we give an insight into what’s going on in the world of tech in various parts of Europe, because, you know, each region has its own strengths, its own weaknesses. And more recently, with the sort of explosion of AI, we’ve spoken about so much. So there’s some interesting things happening in and around the European tech scene as well.
Tom Chitty
Do you think it’s a time where it seems very excitable?
Arjun Kharpal
The founders and CEOs I’ve spoken, I’ve always asked them this question: Europe pretty much lost out in the internet age — the big U.S. tech giants — is this the time Europe gets a chance to catch up? And they’re all very, of course, as you’d expect, very excited. This is a chance Europe can pull out some pretty big companies. But you know, that’s the vibe that I’m getting right now.
Tom Chitty
Okay, let me introduce our special guest. Sanjot Malhi. Malhi is a partner at venture capital fund Northzone. He’s been there for a year and a half. And I’m gonna get you, Sanjot, to explain what Northzone specializes in, where its key focus is. But first, I want to find out about your former career as a professional athlete having played football professionally, as well as club level cricket. Sanjot, talk us through it.
Sanjot Malhi
So I mean, most of my childhood was centered around sports. Not around technology or anything of the sort. So yeah, I played club cricket. So I grew up in the Netherlands. I was born in India. I played club cricket. I grew up in The Hague. So I was captain of the Hague Cricket Club. The Netherlands, as you know, is not a cricket major, so it’s not such a big deal. But then I played football the most.
Tom Chitty
Okay, before we get into it, we have, of course, got to do stat of the week. It’s not that complicated. Arjun will come up with a stat. And we have to try and guess what that stat refers to? It will be related to what we’re talking about in this episode.
Arjun Kharpal
£953.7 million. That’s the stat this week.
Tom Chitty
Okay, Sanjot, first and easy question to start off. If you could rank Europe’s tech hubs, what would be your top three?
Sanjot Malhi
Yeah, so I tend to be quite numbers driven. So my top three would be the U.K., Germany and France, in probably in that order. And the reason is, if you look at 2023’s venture funding landscape, the U.K. was around 22, 23 billion [dollars] in VC funding and the next three combined in Europe are about equal to that and that’s Germany, France and Sweden in that order.
Tom Chitty
I’ve read in the last couple of days that France is coming to take the U.K.’s crown. Is that just hyperbole? Is that scaremongering for us?
Sanjot Malhi
Look, I don’t think it’s a zero-sum game. Zero-sum headlines are great because they get eyeballs. If you pit people against each other; it doesn’t need to be, right? Innovation is unlimited if you do it right. So I don’t think crowns have to be taken from one another. I think both can be great tech hubs. And they hopefully will be if you look at the early signs.
Tom Chitty
And we’ve talked a little bit about where the U.K. is within Europe. Sounds like it’s No. 1. Where is the U.K. globally, though? Because I imagine it’s not No. 1.
Sanjot Malhi
Yeah, no, it’s not. Just to be blunt. Yeah. So I’ve had the pleasure of now, having worked in investment across North America, Asia and Europe. If you look at the great tech hubs around the world, San Francisco, Beijing are up there. Just, if you look at the 1,200, 1,300 unicorns around the world, the vast majority are between those two cities, really. So it’s very, very concentrated. And then you have Tel Aviv, Bangalore, southeast Asia, and Latin Am now coming up as well. And Europe, of course. Europe has about 160-odd unicorns, and the biggest hub within that being the U.K. with 50, or 60. So to answer your question, no, it’s not the biggest in the world. But it’s not insignificant, right? If you look at it, and things are trending, broadly speaking, in the right direction, if you look at the capital, the quality of the founders and really, even, outcomes in terms of companies being created
Arjun Kharpal
Sanjot, just back to the U.K., you know, in the 2010s, mid 2010s, it’s all about fintech. People said London, in particular, is a financial hub, it’s got that sort of history. On top of that, you know, it’s got good pools of capital. At that time, it was part of the European Union, so access to the single market. What now as you look at it in 2024 are the U.K.’s, more broadly, strengths in tech?
Sanjot Malhi
It’s a good question. I think most tech hubs tend to develop around industries where they have incumbent advantages. And London for obvious reasons, has had that in financial services, like you said, so it’s not surprising that fintech was the first sort of starting round for tech activity. But you know, going back to my earlier point about global hubs, if you look at most of these global hubs, they develop in similar manners. So the first sort of wave of innovation comes around whatever is most natural, whatever has been done for the longest time here, it’s financial services. But once you have these large companies here, you know, the likes of Klarna, and Monzo, and others that have come out of that wave, you get a second generation of entrepreneurs, who then tend to solve issues that they faced, because now they’ve grown up in these tech environments. And so they’re facing very different issues to the first generation. And so I think, going forward, you’ll see much more tech-heavy, deeper-tech kind of problems being solved. And that will probably reflect in that makeup of the verticals.
Arjun Kharpal
And what’s happening with venture funding here? Is the recovery taking hold in the early part of 2024 in the first half at the moment, or is there still a bit of caution, given some of the broader macroeconomic headwinds? Of course, in the U.K. we have an election upcoming too?
Sanjot Malhi
So I would say 2022 was probably the worst of the cycle, where we kind of peaked. I mean, I was previously at a firm that did both public and private investing, so you know, it was a good vantage point to see both sides of the equation and things really went downhill at the end of ’21 and early ’22. And it stayed pretty bleak for most of that year. But if you go back to ’23, we’re already seeing growth in terms of funding, which is very good to see. And you’ve seen early signs of the recovery. I think the U.K. is up 30, 40% year-on-year, France is up the most, to your to your point, almost 50%. So I think there are early signs of recovery in ’24. The numbers remain to be seen, but anecdotally, you feel it when you look around; big, big rounds happening, the likes Wayve, even in the U.K. So I think there’s early signs of it and we’ve had an IPO today, which which is kind of a big milestone. So I think things are trending in the right direction.
Arjun Kharpal
What are the thematic, then, that are prominent at the moment for you, as you look at the investment landscape in the U.K.?
Sanjot Malhi
Yeah. So we we tend to be [a] very, very bottoms up sort of fundamentals-driven investor, especially on the later side, which is what I do. So we look for real problems being solved, frankly. And if you look at the big trends, AI is a big one. But for me, AI only matters if you’re solving real problems. AI for the sake of AI is not very interesting. But it will remain a big, big trend. And then, of course, you have software as always, but I think we’re increasingly seeing a lot more hardware, as well, in sort of real-world problems. And I’ve lately been seeing even quantum, especially in the U.K., which is really fascinating.
Tom Chitty
When you talk about 2022, obviously, being a bad year, let’s go even further back to 2016 and Brexit. How has if at all, Brexit, affected the U.K. tech scene?
Sanjot Malhi
Yeah, so if you think about what creates a tech hub? I think the most important and fundamental thing is talent. And then there’s capital. And then there’s you know, regulation and so on, things that flow from there, but really, it’s about the talent, that is where it starts. And so I think the biggest thing it’s affected is the inflow of high-quality talent. There are bigger barriers to people coming in and working in the U.K., because of Brexit. I think the U.K. has had very good policies in terms of, you know, visas being more available and streamlining visa processes, especially for tech workers and so on. But I do think that there’s a lot more that can be done to sort of streamline that and make it more accessible, and therefore, you know, a more thriving sort of tech environment.
Arjun Kharpal
Are there other challenges at the moment, as you see them to the U.K. tech landscape? Is it around, for example, some of their lingering issues from Brexit? Or is it around government policy or anything like that?
Sanjot Malhi
Yeah, this is probably a fairly boring answer. But I think one of the best things a great government can do is create just an enabling environment for innovation and then get out of the way. Because sometimes, one mistakes activity for productivity, and there’s almost a political need to be active and create regulation when it’s not necessary. So I think it’s important to protect the public. And so to that extent, regulation is important. But beyond that, I think it’s really about creating an environment that lowers the hurdles for talent, capital, inflow, company formation, failure, you know, clear bankruptcy laws and things like that. So that, sort of, virtuous cycle that enables people to start up and create new companies, I think, is ultimately what is required.
Arjun Kharpal
So are you suggesting at this point [it] sounds like there’s a bit too much regulation in certain areas in the U.K.? … Because I remember … going back to those days when we’re talking about the fintech boom in London, the FCA at that point was seen as quite a forward-thinking regulator, they created the sandboxes and various other things that allowed firms to experiment with a lot of financial technology products. And they were quite well regarded. I haven’t necessarily heard the same about the modern-day FCA. And also, you know, there’s a lot of criticism of the Competition and Markets Authority and how aggressive they’ve been. So when you talk about the regulatory front, when you talk on that, is there something at the moment in the U.K. that could be improved? Or you feel that’s too burdensome?
Sanjot Malhi
Yeah. Look, I mean, I’m really no expert in regulation. So I’ll refrain from commenting on specific sorts of political policies. But generally speaking, if you look at Europe, it has a history of creating more onerous regulation than some of the other tech hubs, namely the U.S., you know, which is far more, sort of, laissez-faire in many ways, which enables innovation. So that’s, I guess, where my point is coming from, and I think the U.K. has traditionally been part of that European ecosystem. I think governments have gotten that, including the present government, and recognize the need to create a more streamlined environment. But if I’m being honest, the reaction has been missing behind that sort of, you know, political fervor, if you will. We haven’t seen enough to create that environment. I think both parties have already announced that they will be, they’ll have an AI policy. I still don’t know what that means. But you know, [it] remains to be seen what that entails.
Tom Chitty
We’re recording this on Tuesday, the 11th of June, just to be clear. And this morning, we had quite a big IPO, which you covered extensively, Arjun. When it comes to IPOs, in the U.K., [it’s] pretty underwhelming of late. How big was that? Arjun talk through the listeners what happened this morning.
Arjun Kharpal
I walked into the London Stock Exchange, there was confetti everywhere. The team at Raspberry Pi, this is the company you’re referring to, a British computing startup been around since 2012, selling these single-board computers, initially started for hobbyists really. Now they say their biggest market is industrial uses. It’s a pretty small IPO, when you look at tech IPOs more broadly, particularly if you look across the pond to the U.S., you know, it’s a pretty small IPO, I think the valuation in the end was just north of £500 million, so it hasn’t even hit a billion dollars. But it was obviously a big deal for London, which as you mentioned, has just had a dearth of IPOs, and particularly in the tech sector. So there was a lot of excitement around it, the CEO of the London Stock Exchange was around all the executives. So there’s a lot of excitement, [a] big party. I was on the roof. I was on the roof, had a great view.
Tom Chitty
But you weren’t under the confetti?
Arjun Kharpal
No, I heard it pop off. I wasn’t allowed in that area.
Tom Chitty
So I suppose, just talking about that, I mean, is it something that we should be taking note of? Is it a huge boost? Or should we just temper our excitement?
Sanjot Malhi
Yeah, so I think it’s more symbolic than anything else. Like Arjun mentioned, it’s not big. If a unicorn were to go public, it’s not even hit that scale. So in the grand scheme of things, it’s [a] fairly mid-sized to small IPO. But it is symbolic, it is tech, it is in London, and we haven’t seen many of those. If you think about IPOs and capital markets, they are, in some ways, the ultimate network effect businesses, right? It’s a little bit of a cold start problem, you need high-quality businesses to want to list. You need deep-pocketed investors who can deploy large sums of permanent capital. And then, third, you need really intelligent investors in those specific sectors, in this case, tech, that understand the nuances that the U.K. has been missing all three. And therein lies the issue.
Tom Chitty
And how does it reverse that then?
Sanjot Malhi
Yeah, and that’s why, you know, I said, it’s a little bit of a cold start problem. It has to start somewhere. I reckon it’ll start with somebody that’s large, a company choosing to go public here for whatever reason —that might be political, it might be something else. And that kick starts other things in that ecosystem. Because the big mutual funds are all here, fidelity is here. And Franklin Templeton, they’re all here. It’s just that they haven’t had an opportunity to deploy large sums of capital and buy high quality companies. Maybe it could be one of the Chinese, like Shein listing here, and that could potentially be a good effect.
Arjun Kharpal
I have a ton of questions. First one, I mean, how big a snub was the Arm IPO in the end? You know, this is a British company. There’s questions whether the British government back in 2016 should have even allowed this company to be sold. I was in the room at the press conference at the time, [Softbank founder] Masayoshi Son was there. This is one of the most critical semiconductor firms in the world. And now they listed in the U.S., that was a big blow, right?
Sanjot Malhi
I mean, it’s a blow but it’s also somewhat symptomatic of what I just said, right? So it doesn’t necessarily surprise me, also it is Japanese owned right? It’s not really U.K. owned. It may have been started here but it’s owned by Masa and Masa made the call to list in in the U.S. and perhaps with good reason because, you know, like I said, the access to that quality of capital doesn’t exist. But I think somebody has to make that leap. And then the rest follows. It may not be Arm because again, Softbank has its own problems. So you know, they’re protecting their downside more than anything else at that point. And so they were taking the safer option, but it will take somebody to, sort of, make that leap. And I think somebody will.
Arjun Kharpal
I know talking about rules is boring, but sometimes it’s important. The London Stock Exchange in particular, and the regulation around listing in the U.K., has been criticized quite heavily by tech companies around things like foundership, dual-class shares. One of the things that struck me this morning when I was covering the IPO was today, Tuesday, 11th of June 2024, is [the] conditional listing of Raspberry Pi. So only institutional investors effectively can buy up shares. And then it goes on to properly trade three days later on Friday. For the retail investors at this point, we had a 30% pop on the price, the retail investors are sitting there going well, I’ve missed out that whole chunk if it continues to go up. Those kinds of things, I think, really grate tech founders and CEOs of larger companies when they look to list here. So it’s an issue with a whole reform of listing rules here in the U.K., which needs to be carried out, right?
Sanjot Malhi
I think it is a bit of that. Most stock exchanges lag behind the U.S. in terms of listing rules, to your point, the U.S. is definitely the most evolved, they’re more accommodating of sort of venture-funded companies, if you will. Again, going back to the global landscape, the same used to be said of China if you went back a decade, the same used to be said of India five years ago, literally the same thing. And now, people are talking about non-Indian companies listing in India, because that stock exchange is so, so valuable and so attractive. What changed? One, the Indian government said, loss-making companies can now list which was never the case. And then second, somebody just made the leap and said, you know, we’re gonna go list and then the rest followed. I think it’ll be a bit of both in the U.K. as well, where it will take a government that changes a couple of the big things that you mentioned, and then you know, somebody listing, and the rest will fall in place.
Tom Chitty
Startups in the U.K., we hear a lot that the environment isn’t particularly supportive, at least from the U.K. government, and that they could be doing more, that’s obviously their take on it. Where do you see the U.K. government on that front?
Sanjot Malhi
Yeah, I think it goes back to, sort of, the people aspect, I think that is probably the biggest hurdle here in terms of creating employment. So now, the U.K. startup scene employs about 1.8 million people. So it’s not immaterial, it’s significant. And that’s gone up about 4x since 2018. So the growth is really material. People used to say that, you know, there’s no vote bank in tech and so politicians didn’t care about tech as much. I don’t think that’s the case anymore. So I think people, the sort of political class, is taking notice of that. And to your point, I think what can specifically change is employability laws, the ability to attract more talent, retain more talent, make it much more streamlined, to create option pools, award people with equity and the treatment of that equity should be at par with public equities, all of those things, I think, have to fall in place for that ecosystem to work. So in short, I think it’s really about the people aspect more than anything else.
Arjun Kharpal
Sanjay … I want to talk about government’s role. One thing that struck me when I visited France, when you look at Viva Tech. It’s a big show, we know that, but the French government ministers and even [Emmanuel] Macron himself, the president, is in attendance. Just this past few weeks ago, he gathered a bunch of CEOs, from technology companies, and leaders together. And whilst a lot of this, you know, can be seen to be perhaps showboating, there is a genuine feeling that actually the government is throwing its full weight behind the force of France’s tech sector. I don’t think it’s any surprise, you’ve got companies like Mistral AI … etc, raising large sums and being kind of thrust into the spotlight. The U.K. just again, looking from the outside, feels like it wants to talk about tech, it wants to say we’re a leader in crypto, we’re a leader in AI, in all these areas, but the action and that kind of level of support, hasn’t felt [like it’s] there. Is that an accurate reflection in your point of view? Does the U.K. government need to be doing more in a similar vein to what the French are doing at the moment?
Sanjot Malhi
So I think the correlation is certainly that if you have a tech-forward leader, who’s the head of state or head of government, I think it certainly helps, right? It’s been the case around the world again, wherever you look, China, India, the U.S. and Europe now in France, and I think Emmanuel Macron is definitely somebody who is very tech forward and forward leaning on technology. It feels like the prime minister here is well, there’s obviously different sort of, I guess, political complexities in every system that hold different leaders back, but ultimately you want a tech ecosystem that is removed from politics, right? So no single individual, even if they happen to be the head of state should be able to move the needle that much ideally speaking, right? In the U.S., if you look at San Francisco, [it] doesn’t matter who the president is. They have built AI, they have all sorts of innovation, they have all funding flowing. And that’s the ultimate state that you want to arrive at. But yeah, I think France has done a really great job. But at the same time it is today, it’s still very, very early days, right? I mean, still there’s only one company that everyone mentions, when you think of it. It’s just Mistral that everyone’s talking about. So the repeatability of that remains to be seen. And if you look, markets are down in France today, because of the European election and the results there as well. So you don’t know what happens. But ultimately, you want a tech ecosystem that is removed from politics. But yet, you need a government that is supported throughout.
Arjun Kharpal
Yet it’s so hard these days, tech and politics are intertwined. The battle between China and the U.S. over various technologies from AI to semiconductors, it’s very difficult these days for that to happen. And it’s almost as if governments need to, or feel the need to, be involved, in particular areas, I guess, where they see strategic importance, whether it’s around semiconductors, military applications, AI, right? And so it’s those areas they’re getting involved in, but you don’t necessarily want them to be involved in some of these other areas.
Sanjot Malhi
Yeah, no, I completely agree. I think the strategic elements remain and will always remain, right. And … some of the geopolitics is becoming more complicated with the Middle East, and so on and Ukraine. So that is understandable. But for the rest of it, it shouldn’t require so much of the government’s heavy lifting to do it. Ideally, what the government should create is fostering the innovation, which is education, the laws that enable company creation and bankruptcy, they encourage FDI those are the fundamental things, the boring stuff that the government needs to do. But, you know, I think the problem with democracy, which we all love, is that it is prone to populism, right? And so they tend to go toward what makes for the best headlines. And, you know, FDI rules don’t really make for great headlines.
Tom Chitty
Last question from me. Looking ahead now for the U.K. tech scene, where do you think they could hang their jacket? That’s the area they should be aiming to be a real leader in?
Sanjot Malhi
I think several things. AI definitely. Again, going back to the education hubs, I think the U.K. has amongst the best technical education institutes in the world. So AI and software in general, for sure. I think forward leaning, we will see quantum create a massive leap. And I think the U.K. will be a very, very important hub in that and we’re seeing early signs of that. And then even hardware because of that sort of deep tech understanding and education and even climate tech. I would say those four are probably the ones that come to mind.
Arjun Kharpal
And just the last one for me, just because you mentioned AI and you know that’s the topic du jour. Yeah, DeepMind is the company everyone talks about when they talk about U.K.’s AI scene, and it’s obviously owned by an American firm, Alphabet. Does the U.K. have what it takes, in your view to produce a global AI giant?
Sanjot Malhi
Yeah, I mean, I hope so because it affects my life. But look, I mean, we’re in the infancy of AI, right. And we forget that. It’s literally day zero of what the new version of generative AI looks like. And this is somewhat reductive, if you break it up into sort of the four big buckets of AI: No. 1 is hardware, closest to the metal, which is the chips and we’ve seen Nvidia rally. The second is the large language models, which is the Mistrals and Llamas of the world. Then I think there will be an infrastructure software layer built around that, the likes of Snowflake for specifically AI. And then finally, perhaps most importantly, the application layer, because we all speak about AI but it’s not as prevalent in our lives and businesses as it should be yet so there will be an application layer developing. The reason I’m saying all of this is because we are so focused on step one and two today because that’s all we see. Steps three and four, I think, is where the ultimate value capture will lie. And I think the U.K. has as good a chance as any to build something.
Tom Chitty
Okay, fantastic. Well, we’ll leave it there. But before we finish, we have, of course, got to do stat of the week. Have you been thinking about it? Because you seem very focused on the podcast, but you should have really been thinking about stat of the week.
Arjun Kharpal
I think he knows the answer. I think your time’s done. £953.7 million is the stat.
Tom Chitty
I’ll go first because if I’m gonna be wrong, then I don’t want to copy your answer. I’m gonna go with the amount of VC funding in U.K. tech startups for 2025.
Sanjot Malhi
I would say that is the amount of funding in U.K. AI startups.
Arjun Kharpal
That’s wrong. Little hint. It relates to last year. And it relates, I’ll just say, that the phrase Raspberry Pi, on the rooftop of the London Stock Exchange.
Sanjot Malhi
Money raised from IPOs last year.
Arjun Kharpal
There we go.
Tom Chitty
I did concede, I was about to say the answer.
Sanjot Malhi
Too slow, too slow.
Arjun Kharpal
It was the amount raised via IPOs in the U.K. in 2023, down 40% from 2022, in which issuers raised £1.6 billion.
Tom Chitty
Sanjot, thank you very much for joining us on Beyond the Valley. That’s it for this episode. But before we go, please follow and subscribe to the show and rate us if you’d like.
You may like
Technology
AI research takes a backseat to profits as Silicon Valley prioritizes products over safety, experts say
Published
5 days agoon
May 14, 2025By
admin
Sam Altman, co-founder and CEO of OpenAI and co-founder of Tools for Humanity, participates remotely in a discussion on the sidelines of the IMF/World Bank Spring Meetings in Washington, D.C., April 24, 2025.
Brendan Smialowski | AFP | Getty Images
Not long ago, Silicon Valley was where the world’s leading artificial intelligence experts went to perform cutting-edge research.
Meta, Google and OpenAI opened their wallets for top talent, giving researchers staff, computing power and plenty of flexibility. With the support of their employers, the researchers published high-quality academic papers, openly sharing their breakthroughs with peers in academia and at rival companies.
But that era has ended. Now, experts say, AI is all about the product.
Since OpenAI released ChatGPT in late 2022, the tech industry has shifted its focus to building consumer-ready AI services, in many cases prioritizing commercialization over research, AI researchers and experts in the field told CNBC. The profit potential is massive — some analysts predict $1 trillion in annual revenue by 2028. The prospective repercussions terrify the corner of the AI universe concerned about safety, industry experts said, particularly as leading players pursue artificial general intelligence, or AGI, which is technology that rivals or exceeds human intelligence.
In the race to stay competitive, tech companies are taking an increasing number of shortcuts when it comes to the rigorous safety testing of their AI models before they are released to the public, industry experts told CNBC.
James White, chief technology officer at cybersecurity startup CalypsoAI, said newer models are sacrificing security for quality, that is, better responses by the AI chatbots. That means they’re less likely to reject malicious kinds of prompts that could cause them to reveal ways to build bombs or sensitive information that hackers could exploit, White said.
“The models are getting better, but they’re also more likely to be good at bad stuff,” said White, whose company performs safety and security audits of popular models from Meta, Google, OpenAI and other companies. “It’s easier to trick them to do bad stuff.”
The changes are readily apparent at Meta and Alphabet, which have deprioritized their AI research labs, experts say. At Facebook’s parent company, the Fundamental Artificial Intelligence Research, or FAIR, unit has been sidelined by Meta GenAI, according to current and former employees. And at Alphabet, the research group Google Brain is now part of DeepMind, the division that leads development of AI products at the tech company.
CNBC spoke with more than a dozen AI professionals in Silicon Valley who collectively tell the story of a dramatic shift in the industry away from research and toward revenue-generating products. Some are former employees at the companies with direct knowledge of what they say is the prioritization of building new AI products at the expense of research and safety checks. They say employees face intensifying development timelines, reinforcing the idea that they can’t afford to fall behind when it comes to getting new models and products to market. Some of the people asked not to be named because they weren’t authorized to speak publicly on the matter.
Mark Zuckerberg, CEO of Meta Platforms, during the Meta Connect event in Menlo Park, California, on Sept. 25, 2024.
David Paul Morris | Bloomberg | Getty Images
Meta’s AI evolution
When Joelle Pineau, a Meta vice president and the head of the company’s FAIR division, announced in April that she would be leaving her post, many former employees said they weren’t surprised. They said they viewed it as solidifying the company’s move away from AI research and toward prioritizing developing practical products.
“Today, as the world undergoes significant change, as the race for AI accelerates, and as Meta prepares for its next chapter, it is time to create space for others to pursue the work,” Pineau wrote on LinkedIn, adding that she will formally leave the company May 30.
Pineau began leading FAIR in 2023. The unit was established a decade earlier to work on difficult computer science problems typically tackled by academia. Yann LeCun, one of the godfathers of modern AI, initially oversaw the project, and instilled the research methodologies he learned from his time at the pioneering AT&T Bell Laboratories, according to several former employees at Meta. Small research teams could work on a variety of bleeding-edge projects that may or may not pan out.
The shift began when Meta laid off 21,000 employees, or nearly a quarter of its workforce, starting in late 2022. CEO Mark Zuckerberg kicked off 2023 by calling it the “year of efficiency.” FAIR researchers, as part of the cost-cutting measures, were directed to work more closely with product teams, several former employees said.
Two months before Pineau’s announcement, one of FAIR’s directors, Kim Hazelwood, left the company, two people familiar with the matter said. Hazelwood helped oversee FAIR’s NextSys unit, which manages computing resources for FAIR researchers. Her role was eliminated as part of Meta’s plan to cut 5% of its workforce, the people said.
Joelle Pineau of Meta speaks at the Advancing Sustainable Development through Safe, Secure, and Trustworthy AI event at Grand Central Terminal in New York, Sept. 23, 2024.
Bryan R. Smith | Via Reuters
OpenAI’s 2022 launch of ChatGPT caught Meta off guard, creating a sense of urgency to pour more resources into large language models, or LLMs, that were captivating the tech industry, the people said.
In 2023, Meta began heavily pushing its freely available and open-source Llama family of AI models to compete with OpenAI, Google and others.
With Zuckerberg and other executives convinced that LLMs were game-changing technologies, management had less incentive to let FAIR researchers work on far-flung projects, several former employees said. That meant deprioritizing research that could be viewed as having no impact on Meta’s core business, such as FAIR’s previous health care-related research into using AI to improve drug therapies.
Since 2024, Meta Chief Product Officer Chris Cox has been overseeing FAIR as a way to bridge the gap between research and the product-focused GenAI group, people familiar with the matter said. The GenAI unit oversees the Llama family of AI models and the Meta AI digital assistant, the two most important pillars of Meta’s AI strategy.
Under Cox, the GenAI unit has been siphoning more computing resources and team members from FAIR due to its elevated status at Meta, the people said. Many researchers have transferred to GenAI or left the company entirely to launch their own research-focused startups or join rivals, several of the former employees said.
While Zuckerberg has some internal support for pushing the GenAI group to rapidly develop real-world products, there’s also concern among some staffers that Meta is now less able to develop industry-leading breakthroughs that can be derived from experimental work, former employees said. That leaves Meta to chase its rivals.
A high-profile example landed in January, when Chinese lab DeepSeek released its R1 model, catching Meta off guard. The startup claimed it was able to develop a model as capable as its American counterparts but with training at a fraction of the cost.
Meta quickly implemented some of DeepSeek’s innovative techniques for its Llama 4 family of AI models that were released in April, former employees said. The AI research community had a mixed reaction to the smaller versions of Llama 4, but Meta said the biggest and most powerful Llama 4 variant is still being trained.
The company in April also released security and safety tools for developers to use when building apps with Meta’s Llama 4 AI models. These tools help mitigate the chances of Llama 4 unintentionally leaking sensitive information or producing harmful content, Meta said.
“Our commitment to FAIR remains strong,” a Meta spokesperson told CNBC. “Our strategy and plans will not change as a result of recent developments.”
In a statement to CNBC, Pineau said she is enthusiastic about Meta’s overall AI work and strategy.
“There continues to be strong support for exploratory research and FAIR as a distinct organization in Meta,” Pineau said. “The time was simply right for me personally to re-focus my energy before jumping into a new adventure.”
Meta on Thursday named FAIR co-founder Rob Fergus as Pineau’s replacement. Fergus will return to the company to serve as a director at Meta and head of FAIR, according to his LinkedIn profile. He was most recently a research director at Google DeepMind.
“Meta’s commitment to FAIR and long term research remains unwavering,” Fergus said in a LinkedIn post. “We’re working towards building human-level experiences that transform the way we interact with technology and are dedicated to leading and advancing AI research.”
Demis Hassabis, co-founder and CEO of Google DeepMind, attends the Artificial Intelligence Action Summit at the Grand Palais in Paris, Feb. 10, 2025.
Benoit Tessier | Reuters
Google ‘can’t keep building nanny products’
Google released its latest and most powerful AI model, Gemini 2.5, in March. The company described it as “our most intelligent AI model,” and wrote in a March 25 blog post that its new models are “capable of reasoning through their thoughts before responding, resulting in enhanced performance and improved accuracy.”
For weeks, Gemini 2.5 was missing a model card, meaning Google did not share information about how the AI model worked or its limitations and potential dangers upon its release.
Model cards are a common tool for AI transparency.
A Google website compares model cards to food nutrition labels: They outline “the key facts about a model in a clear, digestible format,” the website says.
“By making this information easy to access, model cards support responsible AI development and the adoption of robust, industry-wide standards for broad transparency and evaluation practices,” the website says.
Google wrote in an April 2 blog post that it evaluates its “most advanced models, such as Gemini, for potential dangerous capabilities prior to their release.” Google later updated the blog to remove the words “prior to their release.”
Without a model card for Gemini 2.5, the public had no way of knowing which safety evaluations were conducted or whether DeepMind checked for dangerous capabilities at all.
In response to CNBC’s inquiry on April 2 about Gemini 2.5’s missing model card, a Google spokesperson said that a “tech report with additional safety information and model cards are forthcoming.” Google published an incomplete model card on April 16 and updated it on April 28, more than a month after the AI model’s release, to include information about Gemini 2.5’s “dangerous capability evaluations.”
Those assessments are important for gauging the safety of a model — whether people can use the models to learn how to build chemical or nuclear weapons or hack into important systems. These checks also determine whether a model is capable of autonomously replicating itself, which could lead to a company losing control of it. Running tests for those capabilities requires more time and resources than simple, automated safety evaluations, according to industry experts.
Google co-founder Sergey Brin
Kelly Sullivan | Getty Images Entertainment | Getty Images
The Financial Times in March reported that Google DeepMind CEO Demis Hassabis had installed a more rigorous vetting process for internal research papers to be published. The clampdown at Google is particularly notable because the company’s “Transformers” technology gained recognition across Silicon Valley through that type of shared research. Transformers were critical to OpenAI’s development of ChatGPT and the rise of generative AI.
Google co-founder Sergey Brin told staffers at DeepMind and Gemini in February that competition has accelerated and “the final race to AGI is afoot,” according to a memo viewed by CNBC. “We have all the ingredients to win this race but we are going to have to turbocharge our efforts,” he said in the memo.
Brin said in the memo that Google has to speed up the process of testing AI models, as the company needs “lots of ideas that we can test quickly.”
“We need real wins that scale,” Brin wrote.
In his memo, Brin also wrote that the company’s methods have “a habit of minor tweaking and overfitting” products for evaluations and “sniping” the products at checkpoints. He said employees need to build “capable products” and to “trust our users” more.
“We can’t keep building nanny products,” Brin wrote. “Our products are overrun with filters and punts of various kinds.”
A Google spokesperson told CNBC that the company has always been committed to advancing AI responsibly.
“We continue to do that through the safe development and deployment of our technology, and research contributions to the broader ecosystem,” the spokesperson said.
Sam Altman, CEO of OpenAI, is seen through glass during an event on the sidelines of the Artificial Intelligence Action Summit in Paris, Feb. 11, 2025.
Aurelien Morissard | Via Reuters
OpenAI’s rush through safety testing
The debate of product versus research is at the center of OpenAI’s existence. The company was founded as a nonprofit research lab in 2015 and is now in the midst of a contentious effort to transform into a for-profit entity.
That’s the direction co-founder and CEO Sam Altman has been pushing toward for years. On May 5, though, OpenAI bowed to pressure from civic leaders and former employees, announcing that its nonprofit would retain control of the company even as it restructures into a public benefit corporation.
Nisan Stiennon worked at OpenAI from 2018 to 2020 and was among a group of former employees urging California and Delaware not to approve OpenAI’s restructuring effort. “OpenAI may one day build technology that could get us all killed,” Stiennon wrote in a statement in April. “It is to OpenAI’s credit that it’s controlled by a nonprofit with a duty to humanity.”
But even with the nonprofit maintaining control and majority ownership, OpenAI is speedily working to commercialize products as competition heats up in generative AI. And it may have rushed the rollout of its o1 reasoning model last year, according to some portions of its model card.
Results of the model’s “preparedness evaluations,” the tests OpenAI runs to assess an AI model’s dangerous capabilities and other risks, were based on earlier versions of o1. They had not been run on the final version of the model, according to its model card, which is publicly available.
Johannes Heidecke, OpenAI’s head of safety systems, told CNBC in an interview that the company ran its preparedness evaluations on near-final versions of the o1 model. Minor variations to the model that took place after those tests wouldn’t have contributed to significant jumps in its intelligence or reasoning and thus wouldn’t require additional evaluations, he said. Still, Heidecke acknowledged that OpenAI missed an opportunity to more clearly explain the difference.
OpenAI’s newest reasoning model, o3, released in April, seems to hallucinate more than twice as often as o1, according to the model card. When an AI model hallucinates, it produces falsehoods or illogical information.
OpenAI has also been criticized for reportedly slashing safety testing times from months to days and for omitting the requirement to safety test fine-tuned models in its latest “Preparedness Framework.”
Heidecke said OpenAI has decreased the time needed for safety testing because the company has improved its testing effectiveness and efficiency. A company spokesperson said OpenAI has allocated more AI infrastructure and personnel to its safety testing, and has increased resources for paying experts and growing its network of external testers.
In April, the company shipped GPT-4.1, one of its new models, without a safety report, as the model was not designated by OpenAI as a “frontier model,” which is a term used by the tech industry to refer to a bleeding-edge, large-scale AI model.
But one of those small revisions caused a big wave in April. Within days of updating its GPT-4o model, OpenAI rolled back the changes after screenshots of overly flattering responses to ChatGPT users went viral online. OpenAI said in a blog post explaining its decision that those types of responses to user inquiries “raise safety concerns — including around issues like mental health, emotional over-reliance, or risky behavior.”
OpenAI said in the blogpost that it opted to release the model even after some expert testers flagged that its behavior “‘felt’ slightly off.”
“In the end, we decided to launch the model due to the positive signals from the users who tried out the model. Unfortunately, this was the wrong call,” OpenAI wrote. “Looking back, the qualitative assessments were hinting at something important, and we should’ve paid closer attention. They were picking up on a blind spot in our other evals and metrics.”
Metr, a company OpenAI partners with to test and evaluate its models for safety, said in a recent blog post that it was given less time to test the o3 and o4-mini models than predecessors.
“Limitations in this evaluation prevent us from making robust capability assessments,” Metr wrote, adding that the tests it did were “conducted in a relatively short time.”
Metr also wrote that it had insufficient access to data that would be important in determining the potential dangers of the two models.
The company said it wasn’t able to access the OpenAI models’ internal reasoning, which is “likely to contain important information for interpreting our results.” However, Metr said, “OpenAI shared helpful information on some of their own evaluation results.”
OpenAI’s spokesperson said the company is piloting secure ways of sharing chains of thought for Metr’s research as well as for other third-party organizations.
Steven Adler, a former safety researcher at OpenAI, told CNBC that safety testing a model before it’s rolled out is no longer enough to safeguard against potential dangers.
“You need to be vigilant before and during training to reduce the chance of creating a very capable, misaligned model in the first place,” Adler said.
He warned that companies such as OpenAI are backed into a corner when they create capable but misaligned models with goals that are different from the ones they intended to build.
“Unfortunately, we don’t yet have strong scientific knowledge for fixing these models — just ways of papering over the behavior,” Adler said.
WATCH: OpenAI closes $40 billion funding round, largest private tech deal on record

Technology
Stock trading app eToro pops 40% in Nasdaq debut after pricing IPO above expected range
Published
5 days agoon
May 14, 2025By
admin
Omar Marques | Sopa Images | Lightrocket | Getty Images
Shares of stock brokerage platform eToro popped in their Nasdaq debut on Wednesday after the company raised almost $310 million in its initial public offering.
The stock opened at $69.69, or 34% above its IPO, pushing its market cap to $5.6 billion. Shares were last up more than 40%.
The Israel-based company sold nearly six million shares at $52 each, above the expected range of $46 to $50. Almost six million additional shares were sold by existing investors. At the IPO price, the company was valued at roughly $4.2 billion.
Wall Street is looking to the Robinhood competitor for signs of renewed interest in IPOs after an extended drought. Many investors saw President Donald Trump’s return to the White House as a catalyst before tariff concerns led companies to delay their plans.
Etoro isn’t the only company attempting to test the waters. Fintech company Chime filed its prospectus with the U.S. Securities and Exchange Commission on Tuesday, while digital physical therapy company Hinge Health kickstarted its IPO roadshow, and said in a filing it aims to raise up to $437 million in its offering.
EToro had previously filed to go public in 2021 through a merger with a special purpose acquisition company, or SPAC, that would have valued it at more than $10 billion. It shelved those plans in 2022 as equity markets nosedived, but remained focused on an eventual IPO.
EToro was founded in 2007 by brothers Yoni and Ronen Assia and David Ring. The company makes money through trading-related fees and nontrading activities such as withdrawals. Net income increased almost thirteenfold last year to $192.4 million from $15.3 million in 2023.
The company has steadily built a growing business in cryptocurrencies. Revenue from crypto assets more than tripled to upward of $12 million in 2024, and one-quarter of its net trading contribution stemmed from crypto last year. That is up from 10% in 2023.
EToro said that for the first quarter, it expects crypto assets to account for 37% of its commission from trading activities, down from 43% a year earlier.
Spark Capital is the company’s biggest outside investor, with 14% control after the offering, followed by BRM Group at 8.7%. CEO Yoni Assia controls 9.3%.
Read more CNBC tech news

Technology
5 new Uber features you should know — including a way to avoid surge pricing
Published
5 days agoon
May 14, 2025By
admin
Travelers walk past a sign pointing toward the Uber ride-share vehicle pickup area at Los Angeles International Airport in Los Angeles on Feb. 8, 2023.
Mario Tama | Getty Images
Uber is giving commuters new ways to travel and cut costs on frequent rides.
The ride-hailing company on Wednesday announced a route share feature on its platform, prepaid ride passes and special deals week for Uber One members at its annual Go-Get showcase.
Uber’s new features come as the company accelerates its leadership position in the ride-sharing market and seeks to offer more affordable alternatives for users. It also follows last week’s first-quarter earnings as Uber swung to a profit but fell short of revenue estimates.
“The goal for us as we build our products is to put people at the center of everything, and right now for us, it means making things a little easier, a little more predictable, and above all, just a little more — or a lot more — affordable,” said Uber CEO Dara Khosrowshahi at the event.
Here are some of the big announcements from the annual product event.
Route Share
Users looking to save money on regular routes and willing to walk a short distance can select a shared ride with up to two other passengers through the new route-share feature.
The prepopulated routes run every 20 minutes along busy areas between 6 a.m. and 10 a.m. and 4 p.m. and 8 p.m. on weekdays. The initial program is slated to kick off in seven cities, including New York, San Francisco, Boston and Chicago.
Source: Uber
Uber said its new route-share fares will cost up to 50% less than an UberX option, and that it is working to partner with employers on qualifying the feature for commuter benefits. Users can book a seat from 7 days to 10 minutes before a pickup departure.
Ride Passes
Riders on Uber can now prepurchase two different types of ride passes to hold fares on frequented routes during a one-hour period every day. For $2.99 a month, riders can buy a price lock pass that holds a price between two locations for one hour every day. The pass expires after 30 days or a savings total of $50.
The feature gives riders a way to avoid surge pricing.
Ride Passes roll out in 10 cities on Wednesday, including Dallas, Orlando and San Francisco, and can be purchased for up to 10 routes a month. Uber will charge users a lower price if the fare is cheaper than the pass at departure time.
The company also debuted a prepaid pass option, allowing users to pay in advance and stock up on regular monthly trips. Uber’s pass option comes in bundles of 5, 10, 15 and 20-ride increments, with corresponding discounts between 5% and 20%.
Both pass options will be available on teen accounts in the fall, Uber said. The route share and ride passes will be available in a new commuter hub feature on the app coming later this year.
Shared autonomous rides
Uber is also expanding its autonomous vehicle partnership with Volkswagen.
The company will start testing shared AV rides later this year and is aiming for a launch in Los Angeles in 2026.
Uber rolled out autonomous rides in Austin, Texas, in March through its agreement with Alphabet-owned Waymo and is preparing for an Atlanta launch this summer. The company announced the partnership in May 2023. Autonomous Waymo rides are also currently offered through the Uber app in Phoenix, but the company does not directly manage that fleet.
Khosrowshahi called AVs “the single greatest opportunity ahead for Uber” during the company’s earnings call last week and said the Austin debut “exceeded” expectations. The company previously had an AV unit that it sold in 2020 as it faced high costs and a series of safety challenges, including a fatal accident.
Along with Volkswagen and Waymo, Uber has joined forces with Avride, May Mobility and self-driving trucking company Aurora for autonomous ride-sharing and freight services in the U.S. The company has partnerships with WeRide, Pony.AI and Momenta internationally.
Uber One Member Days
Uber is taking a page out of Amazon’s book by offering its own variation of the e-commerce giant’s beloved Prime Day, with special offers between May 16 and 23 for Uber One members.
Some of those deals include 50% off shared rides and 20% off Uber Black. The platform is also adding a new benefit of 10% back in Uber credits for users that use Uber Rent or book Lime rides.
UberEats partnership with OpenTable
UberEats also announced a partnership with OpenTable to allow users to book reservations and rides.
The new feature, powered by OpenTable, launches in six countries including the U.S. and Australia.
Through the partnership, users can book restaurant reservations and get a discount on rides. OpenTable members will also be able to transfer points to Uber and UberEats. The company is also offering OpenTable VIPs a six-month free trial of Uber One.
Read more CNBC tech news
Trending
-
Sports3 years ago
‘Storybook stuff’: Inside the night Bryce Harper sent the Phillies to the World Series
-
Sports1 year ago
Story injured on diving stop, exits Red Sox game
-
Sports2 years ago
Game 1 of WS least-watched in recorded history
-
Sports2 years ago
MLB Rank 2023: Ranking baseball’s top 100 players
-
Sports4 years ago
Team Europe easily wins 4th straight Laver Cup
-
Environment2 years ago
Japan and South Korea have a lot at stake in a free and open South China Sea
-
Environment2 years ago
Game-changing Lectric XPedition launched as affordable electric cargo bike
-
Business3 years ago
Bank of England’s extraordinary response to government policy is almost unthinkable | Ed Conway