According to Sifted, 2025 was the year AI agents captured investor frenzy, with Sweden’s Lovable hitting a $6.6 billion valuation. Revolut’s valuation ballooned to $75 billion as it rolled out private banking and reward credit cards to challenge giants like UBS and American Express. The UK government committed £2 billion to an AI sovereignty plan but faced ridicule for offering a “laughable” £80,000 salary for a key role. Meanwhile, climate tech funding plummeted, stalwart tech media like TechCrunch Europe shut down, and the NATO Innovation Fund was rocked by partner departures and alleged conflicts of interest. The year also saw major scoops, including a £6.5 million fraud ruling against The Family’s Oussama Ammar and the launch of Europe’s first VC fundraising accelerator by Mountside Ventures.
Agents everywhere, but where’s the money?
So AI agents were the story. Investors threw record cash at them, and a company like Lovable becomes a near-$7 billion poster child in just 12 months. That’s insane growth. But here’s the thing: we’ve seen this hype cycle before. The article itself notes that “once-hot AI companies began to run into trouble as competition mounted.” That’s the quiet part, right? Everyone’s funding 18 different AI agent startups for every conceivable use case, but how many have a real, durable business model yet? It feels like 2025 was the year of the land grab, and 2026 will be the year of the brutal consolidation. The funding can’t flow like this forever, especially with VC fundraising overall remaining “muted.”
Revolut: the unstoppable question mark
Revolut’s 2025 is a masterclass in aggressive, sprawling expansion. Private banking? Check. Reward credit cards? Check. Mobile plans? Sure, why not. They’re basically trying to be a entire financial ecosystem, and with a $75 billion valuation, the market is betting they can. But the elephant in the room just won’t leave: they still don’t have that full UK banking licence. How long can you be a $75 billion fintech and not have the core regulatory badge of honor in your home market? It’s a massive vulnerability that all the new product launches can’t fully paper over. And the continued silence on IPO plans? It’s starting to feel less like strategic secrecy and more like… they might not be able to.
Ecosystem shocks and sovereign stumbles
The collapse of TechCrunch’s European operations is a bigger deal than it seems on paper. It was a “gut punch” because it’s a signal. When a foundational media pillar for a startup ecosystem vanishes, it makes the whole scene feel less permanent, less supported. It creates an information vacuum. And speaking of vacuums, the UK’s AI sovereignty push is a perfect microcosm of government tech ambition. You pledge £2 billion and bring in smart advisors, which is great. But then you offer a paltry £80k for a leadership role? That’s not even a good salary for a senior engineer at a mid-tier startup, let alone a government AI czar. It shows a fundamental misunderstanding of the market you’re trying to influence. You can’t have “sovereignty” on the cheap.
Drama as a constant
Basically, 2025 proved that European tech is never boring. The NATO Innovation Fund saga—with partner exodus and conflict allegations—reads like a corporate thriller and highlights how hard it is to run a massive, politically-charged VC fund. The VC “mafia” family trees are fun to map, but they also underscore how insular and network-driven the funding scene can be. And the climate tech “pivot to resilience” is just a polite way of saying the easy money dried up and now it’s a brutal fight for survival. So, can 2026 be equally lively? I suspect so. But the liveliness might start feeling a lot more like panic for some sectors that soared a little too high, a little too fast this year.
