ChatGPT

It is believed to be a first: the deliberate targeting of a commercial datacentre by the armed forces of a country at war.

At 4.30am on Sunday morning, an Iranian Shahed 136 drone struck an Amazon Web Services datacentre in the United Arab Emirates, setting off a devastating fire and forcing a shutdown of the power supply. Further damage was inflicted as attempts were made to suppress the flames with water.

Soon after, a second data centre owned by the US tech company was hit. Then a third was said to be in trouble, this time in Bahrain, after an Iranian suicide drone turned to fireball on striking land nearby.

Iranian state TV has claimed that Iran’s Islamic Revolutionary Guard Corps launched the attack “to identify the role of these centres in supporting the enemy’s military and intelligence activities”.

The network built by Jeff Bezos’s company could withstand one of its regional centres being taken out of action but not a second, let alone a third of their huge warehouses of technology.

The coordinated strike had an immediate impact.

Millions of people in Dubai and Abu Dhabi woke up on Monday unable to pay for a taxi, order a food delivery, or check their bank balance on their mobile apps.

Whether there was a military impact is unclear – but the strikes swiftly brought the war directly into the lives of 11 million people in the UAE, nine out of 10 of whom are foreign nationals. Amazon has advised its clients to secure their data away from the region. A technician works at an Amazon Web Services AI data centre A technician works at an Amazon Web Services AI data centre. The drone strikes brought the war directly to 11 million people in the UAE unable to pay for a taxi, order a food delivery, or check their bank balances. Photograph: Noah Berger/Reuters

Perhaps more significantly, the strikes on this ‘next generation’ war target are now raising questions about the prospects of the UAE building on its plans, and many billions of pounds worth of US and other foreign investment, to exploit what they hope will be the ‘new oil’: artificial intelligence (AI).

“The UAE really wants to be a major AI player,” said Chris McGuire, an AI and technology competition expert who served as a White House national security council official in Joe Biden’s administration. “Their government has very strong conviction about this technology, probably stronger than any other government in the world, and if there’s going to start to be security questions around that, then they’re going to have to resolve those very quickly, somehow.”

A datacentre is a facility designed to store, manage, and operate digital data.

The growing demand by businesses for artificial intelligence (AI) and cloud computing – where firms have a pay-as-you-go relationship with the providers of servers, storage and software – is driving the need for centres that have significantly more computational power.

It requires a ready and consistent supply of very cheap electricity.

The UAE, as it seeks to diversify away from fossil fuels, has been able to point out that it has this in spades, along with a huge sovereign wealth fund ready to invest and subsidise projects.

According to Turner & Townsend’s Global Data Centre Index, the overall global cost increase of datacentre construction increased in 2025 by 5.5% – but the UAE ranks 44th in the league table of most expensive unit cost per watt out of 52.

The UAE’s geography also makes it a critical subsea cable landing point, providing access between Europe and Asia.

Then there are the geo-politics, with the US keen to keep the Gulf states away from Chinese technology.

A four-day tour by Donald Trump of Saudi Arabia, Qatar, and the UAE last May coincided with the announcement of the construction of a vast new AI campus – a partnership between the UAE and the US – for the purpose of training powerful AI models.

As part of the deal, the Trump administration eased restrictions on advanced chips sales to the Gulf. OpenAI has said the planned UAE campus could eventually serve half the world’s population. US president, Donald Trump walks alongside United Arab Emirates president, Sheikh Mohamed bin Zayed Al Nahyan The US president, Donald Trump, meets the United Arab Emirates president, Sheikh Mohamed bin Zayed Al Nahyan, in Abu Dhabi during his four-day tour of Gulf states last year. Photograph: Amr Alfiky/Reuters

McGuire said that this week’s events could be pivotal. “If we’re going to have large scale datacentres built out in the Middle East, we’re going have to get pretty serious about how we protect them,” he said. ‘We think about how to protect it right now, and we’re saying, ‘Oh, it means you have guards and good cybersecurity’.

“If you’re actually going to double down the Middle East, maybe it means missile defence on datacentres.”

Sean Gorman, the chief executive of Zephr.xyz, a technology firm that is a contractor to the US air force, said that the Gulf states’ ambitions would have likely been in the thoughts of military planners in Tehran.

He said: “I believe the Iranians are building on tactics they’ve seen be effective in the Ukraine conflict. Asymmetric warfare that can target critical infrastructure creates pressure on adversaries by disrupting public safety and economic activity.

“UAE and Bahrain have both been positioning themselves as global AI hubs by investing heavily in datacentres and fibre infrastructure to connect them to the rest of the world.

“If they can disrupt that infrastructure, it puts their strategic position under risk while also disrupting operations that are important to the economy. In addition, there could be an adjacent impact of defence operations, but that would likely be more luck than the primary objective.”

Gorman said the UAE had a “long track record of managing regional instability without becoming party to it” but that there were a range of risks apart from that from the air.

He said: “The UAE also has one of the most diversified submarine-cable landing environments in the Middle East, but the diversity is geographically uneven.

“There are multiple landing stations and cable systems, but many of them concentrate on the east coast at Fujairah, which creates a partial geographic chokepoint. Attendees at an annual cloud computing conference walk past the Amazon Web Services logo Analysts say operators of prominent datacentres like Amazon Web Services may in future have to invest in air defence, just as maritime operators do against pirates. Photograph: Reuters Staff/Reuters

“In addition, there is a specific risk from Iranian cyber operations targeting US-aligned digital infrastructure in the Gulf, which presents a more concrete near-term threat to datacentre and cloud operations than geography in the traditional sense.”

Gorman said the concern would be if Iran demonstrated any further capability to target Gulf digital infrastructure as part of its retaliation.

He said: “The UAE will need to show partners that its infrastructure is defensible. This is the question investors should be asking, not whether the broader AI ambition survives.”

Vili Lehdonvirta, senior fellow at the Oxford Internet Institute, University of Oxford, said there were significant costs to such defences but that the danger was real.

The former chair of the US National Security Commission on AI, Eric Schmidt, suggested last year that a country falling behind in an AI arms race could bomb their adversary’s datacentres.

Lehdonvirta said he suspected that no one actually believed that datacentres “would get bombed despite such scenarios being openly floated for some time”.

“If that’s the case then from now on we might perhaps see operators of prominent datacentres like AWS [Amazon Web Services] investing in air defence, similar to how shipping operators armed up against pirates,” he said.

Where might Iran fruitfully strike next?

“The Iranians will be well aware that the fibreoptic cables that connect these datacentres to the United States and to the rest of the world run through the strait of Hormuz,” Lehdonvirta said, “although they’ll be closely watched by the US and allied forces.”

In 1987, economist and Nobel laureate Robert Solow made a stark observation about the stalling evolution of the Information Age: Following the advent of transistors, microprocessors, integrated circuits, and memory chips of the 1960s, economists and companies expected these new technologies to disrupt workplaces and result in a surge of productivity. Instead, productivity growth slowed, dropping from 2.9% from 1948 to 1973, to 1.1% after 1973.

Newfangled computers were actually at times producing too much information, generating agonizingly detailed reports and printing them on reams of paper. What had promised to be a boom to workplace productivity was for several years a bust. This unexpected outcome became known as Solow’s productivity paradox, thanks to the economist’s observation of the phenomenon.

“You can see the computer age everywhere but in the productivity statistics,” Solow wrote in a New York Times Book Review article in 1987.

New data on how C-suite executives are—or aren’t—using AI shows history is repeating itself, complicating the similar promises economists and Big Tech founders made about the technology’s impact on the workplace and economy. Despite 374 companies in the S&P 500 mentioning AI in earnings calls—most of which said the technology’s implementation in the firm was entirely positive—according to a Financial Times analysis from September 2024 to 2025, those positive adoptions aren’t being reflected in broader productivity gains.

A study published this month by the National Bureau of Economic Research found that among 6,000 CEOs, chief financial officers, and other executives from firms who responded to various business outlook surveys in the U.S., U.K., Germany, and Australia, the vast majority see little impact from AI on their operations. While about two-thirds of executives reported using AI, that usage amounted to only about 1.5 hours per week, and 25% of respondents reported not using AI in the workplace at all. Nearly 90% of firms said AI has had no impact on employment or productivity over the last three years, the research noted.

However, firms’ expectations of AI’s workplace and economic impact remained substantial: Executives also forecast AI will increase productivity by 1.4% and increase output by 0.8% over the next three years. While firms expected a 0.7% cut to employment over this time period, individual employees surveyed saw a 0.5% increase in employment. Solow strikes back

In 2023, MIT researchers claimed AI implementation could increase a worker’s performance by nearly 40% compared to workers who didn’t use the technology. But emerging data failing to show these promised productivity gains has led economists to wonder when—or if—AI will offer a return on corporate investments, which swelled to more than $250 billion in 2024.

“AI is everywhere except in the incoming macroeconomic data,” Apollo chief economist Torsten Slok wrote in a recent blog post, invoking Solow’s observation from nearly 40 years ago. “Today, you don’t see AI in the employment data, productivity data, or inflation data.”

Slok added that outside of the Magnificent Seven, there are “no signs of AI in profit margins or earnings expectations.”

Slok cited a slew of academic studies on AI and productivity, painting a contradictory picture about the utility of the technology. Last November, the Federal Reserve Bank of St. Louis published in its State of Generative AI Adoption report that it observed a 1.9% increase in excess cumulative productivity growth since the late-2022 introduction of ChatGPT. A 2024 MIT study, however, found a more modest 0.5% increase in productivity over the next decade.

“I don’t think we should belittle 0.5% in 10 years. That’s better than zero,” study author and Nobel laureate Daron Acemoglu said at the time. “But it’s just disappointing relative to the promises that people in the industry and in tech journalism are making.”

Other emerging research can offer reasons why: Workforce solutions firm ManpowerGroup’s 2026 Global Talent Barometer found that across nearly 14,000 workers in 19 countries, workers’ regular AI use increased 13% in 2025, but confidence in the technology’s utility plummeted 18%, indicating persistent distrust.

Nickle LaMoreaux, IBM’s chief human resources officer, said last week the tech giant would triple its number of young hires, suggesting that despite AI’s ability to automate some of the required tasks, displacing entry-level workers would create a dearth of middle managers down the line, endangering the company’s leadership pipeline. The future of AI productivity

To be sure, this productivity pattern could reverse. The IT boom of the 1970s and ’80s eventually gave way to a surge of productivity in the 1990s and early 2000s, including a 1.5% increase in productivity growth from 1995 to 2005 following decades of slump.

Economist and Stanford University’s Digital Economy Lab director Erik Brynjolfsson noted in a Financial Times op-ed the trend may already be reversing. He observed that fourth-quarter GDP was tracking up 3.7%, despite last week’s jobs report revising down job gains to just 181,000, suggesting a productivity surge. His own analysis indicated a U.S. productivity jump of 2.7% last year, which he attributed to a transition from AI investment to reaping the benefits of the technology. Former Pimco CEO and economist Mohamed El-Erian also noted job growth and GDP growth continuing to decouple as a result in part of continued AI adoption, a similar phenomenon that occurred in the 1990s with office automation.

Slok similarly saw the future impact of AI as potentially resembling a “J-curve” of an initial slowdown in performance and results, followed by an exponential surge. He said whether AI’s productivity gains would follow this pattern would depend on the value created by AI.

So far, AI’s path has already diverged from its IT predecessor. Slok noted in the 1980s, an innovator in the IT space had monopoly pricing power until competitors could create similar products. Today, however, AI tools are readily accessible as a result of “fierce competition” between large language model-buildings driving down prices.

Therefore, Slok posited, the future of AI productivity would depend on companies’ interest in taking advantage of the technology and continuing to incorporate it into their workplaces. “In other words, from a macro perspective, the value creation is not the product,” Slok said, “but how generative AI is used and implemented in different sectors in the economy.”

Czech ice dancers Kateřina Mrázková and Daniel Mrázek made their Olympic debut on Monday, an unfathomable feat that takes a lifetime of dedication and practice. But the sibling duo used AI music in their rhythm dance program, which doesn’t break any official rules, but serves as a depressing symbol of how absolutely cooked we are.

As Mrázek spun his sister in a crazy cartwheel-lift-sort-of-move that made them look superhuman, one of the NBC commentators mentioned in passing, “This is AI generated, this first part,” referring to the music. Somehow, that admission is even more baffling than the gravity-defying tricks that the siblings showed off on the pressure of Olympic ice.

The Olympic ice dance competition is split into two events: the rhythm dance, where pairs must perform a routine that meets a specific theme, and the free dance. This season’s theme is “The Music, Dance Styles, and Feeling of the 1990s.” British ice dancing duo Lilah Fear and Lewis Gibson paid tribute to the Spice Girls, while United States favorites Madison Chock and Evan Bates skated to a Lenny Kravitz medley.

But, for whatever reason — licensing issues? — Mrázková and Mrázek danced to a routine with music that’s half AC/DC and half AI. It’s weird. What’s even weirder is that this isn’t the duo’s first use of AI, nor is it the first time that this choice backfired.

Per the International Skating Union, the governing body that oversees competitive ice skating, the duo’s music choice for the rhythm dance this season has been “One Two by AI (of 90s style Bon Jovi)” and “Thunderstruck by AC/DC.” The official Olympics website confirms that the duo is using the AI-generated song for the rhythm dance portion.

@g_nielsenart

ok I know this is an art account but I have been seething about this ever since Shana Bartel caught it and wrote about it in her blog (which you should be reading it’s very good – link below) and I just NEEDED to get it out of me. https://www.patreon.com/posts/142706982 #icedance #icedancing #figureskating #IceSkating #plagiarism
♬ original sound – G Nielsen Art

The Czech siblings have faced backlash before for using AI-generated music. Earlier in the season, they played a ’90s-inspired song for their routine that began with a wailing declaration: “Every night we smash a Mercedes-Benz!” If that sounds familiar, it’s because that lyric comes directly from the ’90s hit “You Get What You Give” by New Radicals (which, by the way, has an incredible music video shot in a Staten Island mall — the true essence of American suburbia!).

The AI-generated lyrics also include the lines, “Wake up, kids/We got the dreamer’s disease,” and “First we run, and then we laugh ’til we cry.” What a coincidence! Those lyrics also appear in the song “You Get What You Give” by New Radicals. The AI song is even titled “One Two,” which are the first words of… you can probably guess which song at this point. TechCrunch Founder Summit 2026: Tickets Live On June 23 in Boston, more than 1,100 founders come together at TechCrunch Founder Summit 2026 for a full day focused on growth, execution, and real-world scaling. Learn from founders and investors who have shaped the industry. Connect with peers navigating similar growth stages. Walk away with tactics you can apply immediately

Save up to $300 on your pass or save up to 30% with group tickets for teams of four or more. Boston, MA | June 23, 2026 REGISTER NOW

Before the Olympics, the duo changed the song, swapping out the New Radicals lyrics for other AI-generated lyrics that sound suspiciously like Bon Jovi lyrics, as journalist Shana Bartels noted in November. For example, “raise your hands, set the night on fire” also appear in “Raise Your Hands” by Bon Jovi… and the AI “vocalist” sounds a lot like Bon Jovi, too. (Not to pour salt on the wound, but “Raise Your Hands” isn’t even from the ’90s!) This was the music that the duo danced to on Monday at the Olympics, before it transitioned into “Thunderstruck” by AC/DC, a real song from the ’90s written by real people.

While it’s unclear what software the team used to generate this music, this is an LLM operating as it’s supposed to. These LLMs are trained on large libraries of music, often through legally dubious means. When prompted, LLMs produce the most statistically probable response to an input. That’s useful when writing code, but means a song “in the style of Bon Jovi,” will likely end up using some actual Bon Jovi lyrics.

And yet, the music industry seems at least temporarily enamored with the idea of “musicians” who aren’t totally real. Telisha Jones, a 31-year-old in Mississippi, used Suno to set her (hopefully real) poetry to music under the persona Xania Monet. Now she has a $3 million record deal.

It’s a shame that these Czech dancers’ accomplishment of skating at the Olympics may be marred by discourse around their use of AI music (discourse that I am actively contributing to). But come on! Isn’t this sport supposed to be creative?

“Vibe coding is a nightmare and I’m getting ready to ban it,” said “Clint,” the CTO of a mid-sized fintech.

He’s not kidding around.

“We opened more security holes in 2025 than we did in all of 2020 to 2024. It’s a miracle we haven’t been breached yet. We keep catching flaws in regression testing – which is pretty late – and at some point we’re going to miss something, and then it’s someone’s head. Probably mine.”

In the back half of last year, I’ve heard from a growing number of tech leaders, and current and former software developers, about their chaotic journey with AI coding in enterprise tech. It was almost always a journey that started with “vibe coding” as a first experimental step.

Now those leaders and developers are implying that vibe coding is just a fad, maybe a marketing ploy to sell AI coding into the enterprise, because if AI allows anyone to be able to code, it triggers the question:

“How many of these expensive software developers do we really need?”

That’s a lot of smoke. I’m a former expensive software developer, a current entrepreneur, and a half-hearted vibe coder. I’m going to recklessly speculate about vibe coding based on some conspiratorial-sounding conversations I’ve had with tech leaders and experienced software developers, and see if there’s fire. My Own Personal Hacker

Yes. Anyone can code. But security holes are easy to open, and thus, “anyone” can also attract their own personal hacker.

Like I did. Twice.

Warning: I am a former developer and, these days, I have a standby network of current senior developers who can help me with things like security. Don’t try this at home.

Without naming names, over the last several years, I built a minor empire out of a no-code platform hooked up to a lot of low-code tools. I did this partly to show that anyone could do this – “Anyone can no-code!” And it worked great and still works today.

Except a few years ago I spent an entire summer fighting one individual, persistent hacker. No flaws to fix, no holes to patch, but I couldn’t stop him from trying to break in. The no-code platform couldn’t stop him. His ISP wouldn’t do anything. Finally, out of frustration, I just ditched the whole function and leaned totally into Stripe. More expensive, and a lot of hours lost, but no more hackers, and thankfully, no damage done.

More recently, I vibe coded my way to a beautiful web-app with a sorta-simple login function to privately self-serve the showing of demos of some of the apps I’ve built, genericized and anonymized—so no sensitive data to be hacked, but the hackers came anyway.

I took that app down too. Now I do demos the old fashioned way, from my local machine over Zoom. It’s sooo 2023. If You Vibe Code It, The Bots Will Come

In both cases, my developer friends were not shocked at my stories, and frankly, by the second time, neither was I. When I asked why these hackers would come after me, they told me “because you’re there,” and that the second “hacker” was almost assuredly a bot or team of bots, just sniffing out public apps and breaking into them.

“If you build it, they will come,” said my favorite CTO Ryan Eade. And what he’s talking about is a recent but rather well-known conundrum. Anyone can code, but the minute they go live, they’re setting themselves up for security failures that these vibe-coding tools just weren’t built to handle.

So knowing this, why were the AI platforms pushing vibe-coding so hard, and why wasn’t anyone screaming about the security nightmare that was unfolding across the internet?

Well, I’m being facetious. They were screaming. Hell, I was too. It’s just that no one was listening because everyone was too busy building the next billion-dollar vibe-coded app.

Or too busy firing senior developers they believed they no longer needed.

Here’s where it gets spicy. The Old Man and the C++

I’ve been wanting to use that awful pun for months now. My editor will love it.

Over the last year-plus as I’ve been documenting the decimation in the tech industry, including, specifically, the gutting of senior talent from tech teams, I’ve collected a following of these former developer cast-offs.

Let’s face it. They’re mostly old men, or slightly older men, and definitely a lot more women than before, but not as many in the 40+ age range. These are the folks with 15-20 years of experience. They were and are the hardest hit by tech disemployment, and they’re pretty sure they know why.

“[They] gave [AI coding] to us,” said “Merlin,” one of the middle-aged former developers who follows me. “They made us prove we should keep our jobs. We did. I did, I know I did. But then they lopped off the top… 60 percent or so of our team. By experience. They claimed ‘streamlining for the future’ but it was all us greybeards that got the ax.”

Merlin is not alone. His is one of dozens of similar stories I’ve heard over the last 18 months. And while there are a lot of reasons – a lot of reasons – for companies to do mass layoffs in tech these days, a number of these folks are now coming around to what in retrospect feels like an exercise in AI coding that was designed to push them out.

“It’s not that the tools were passing us by,” Merlin said. “The tools were pretty cool and I picked them up quickly. It was more about ‘If we have AI, why do we need the programmers?’” The Answer Is Security

Clint is still trying to shove vibe-coding back into the tube it came from. I asked him what he thought about the vibe-coding wave being promoted to push senior talent out in favor of AI coding tools.

“I don’t know,” he said. “I don’t see how anyone is doing this at scale who isn’t completely versed in infrastructure, security, privacy, overall data governance.”

Clint’s company didn’t lop off any heads, but now they’ve got standards in place for when AI coding agents are to be interfaced with and how. He’s frustrated, but at least at this point he’s putting out dying embers, not roaring fires.

I asked him why he didn’t take a more cautious approach at the beginning. He paused for a few seconds.

“OK. Honestly, there were a couple reasons. I’ll admit I was just as excited to get into AI as anyone else. But also, there’s this nagging pressure, like, if we don’t pick this stuff up quickly, we’re going to get left behind by our competitors who do.”

He laughed. “Or some kid in a dorm room who rebuilds our entire business over a weekend.” Then he paused again. “I know that’s not going to happen.” Vibe Coding Wasn’t a Ruse, But It Left a Mark

Of course it won’t. But it’s that fear that drove a lot of companies to dump talent overboard and sink those dollars into an “AI first” infrastructure. And it’s still happening today. A kid in a dorm room isn’t going to replace a 100-person tech team overnight. But why carry that liability when 80 techies will do? Or 50? Or 10?

“There are a lot of us,” Merlin said, referring to the growing number of senior, now unemployed developers on the sidelines. “Maybe a kid can’t rebuild a business in a weekend, but I feel like a few of us could do something like that. It’d be sweet payback to take on these companies with their own strategy.”

Ultimately, I think vibe coding took off on its own, riding the same wave that made me stand up my own apps. There’s no question some tech companies also rode that wave and used it as an excuse to take drastic action to assuage those nagging fears. And as a bonus, they got a better bottom line to boot.

Just remember, in their wake, they left a growing, experienced, motivated army of developers who are much more efficient with these AI coding tools, and they aren’t interested in “vibes” so much as “disruption.”

I’ll talk about some of the more disruptive approaches in AI coding in future posts. Now would be a good time to join my email list, a growing army of professionals who want a unique take on the hype and the histrionics.