With rising talks of an ‘AI Bubble’ – the risk of rapid AI innovation and hype leading to an eventual market crash – there is growing uncertainty on how to approach the fast-growing technology. Is investing in AI worth the risk? Five investors share their thoughts on if there is an AI bubble at all, and what businesses can do to prepare.
Omar Rajjoub, Financial Advisor, HudsonPoint Capital
“Very very few have ever been able to predict a bubble in the past and this would be the first time the masses have ever been aware of and been able to see what’s coming. I do not think we are in a bubble, I think we are in one of the most aggressive growth cycles we have seen in decades. I liken it to our generation’s industrial revolution. There are going to be winners and losers, destruction of capital, the birth of new bellwether businesses and new ways of life.
“I think people are overestimating what AI can do inside the next year or two and can barely comprehend what will be capable in the next decade. Long term, the applications, the use cases are nearly endless. We barely even talk about visual models and haven’t even cracked AGI yet. It is going to change the way we live.
“Only a few of the models are going to survive and generally in these circumstances it is going to be the ones that were first to market, i.e. OpenAI. Virtually all the seed and the pop-up ventures with .ai at the end of their name will be gone. The lucky ones will be acquired before that.
“You cannot avoid capital destruction in a growth cycle. It happened in the dot-com boom and bust. The emergence of the cloud business. Less spoken of however, [is] the shale expansion in the US. The only way to avoid any type of capital destruction is not to play the game. Don’t invest in any companies involved in AI, don’t make any money with them but don’t lose any either.”
Dan Buckley, Chief Stock Market Analyst, DayTrading
“Lots of value is being capitalised in the present in anticipation of future earnings and productivity benefits. I think what’s clear is that there’s going to be a lot of investment that’ll turn out to not provide any return. Personally, I’ve lowered my exposure to US equities where a lot of forward growth has been baked into the valuations, as forward returns are likely to come in lower.
“AI stocks are high at the same time there’s a lot of stimulation from fiscal policy and monetary policy, along with relatively loose regulatory policy. The pure-play AI companies with unclear unit economics that rely heavily on external funding, and especially the names using AI as a story to promote their valuations, are the most at risk. I think the technologies themselves will provide long-term boosts in productivity in certain ways. It’s nonetheless important to recognise that just because a technology, or group of technologies, is innovative it doesn’t necessarily mean that it’s highly profitable to produce.”
Samyr Laine, Managing Partner and GP, Freedom Trail Capital
“We’re in a hype-cycle, not a bubble. The difference matters. AI has real utility, it’s already improving efficiency, personalisation, and decision-making across industries. But we’re seeing inflated valuations for companies slapping “AI-powered” on pitches without demonstrable product-market fit or sustainable unit economics. The froth is in the funding rounds, not the technology itself. In consumer brands specifically, AI is helping founders optimise supply chains, predict trends, and personalise at scale with tangible results.
“The bubble bursts when investors demand ROI and most AI companies can’t deliver. Right now, too much capital is chasing “AI-enabled” businesses with no path to profitability. The survivors will be businesses with strong unit economics, loyal customers, and founders who built sustainable models before layering AI on top. Capital efficiency will separate winners from casualties.
“Investors need to stop rewarding buzzwords and start demanding proof: What’s the cost per customer with and without AI? How does this make your business defensible? Founders need to be honest about what AI actually does for them. If it’s marginal, don’t lead with it. If it’s transformational, show the data. The companies that avoid getting caught in the correction will be the ones that treated AI as infrastructure, not story. Build a profitable business first. Use AI to scale it faster and run leaner. Don’t build an AI company and hope to find a business model later.
“The winners will be brands that use AI to enhance the human experience, not replace it. The future belongs to entrepreneurs who master AI as a tool while staying rooted in what makes their brand authentic.”
Kamran Ansari, Venture Partner, Infinity Ventures
“The real test comes when lockups expire. A handful of fintechs have gone public recently (Klarna, Chime, Etoro, Circle), but they are all still in lockup. The first six months of trading are mostly theater because insiders, founders, and funds cannot sell. The real signal will come in 2026 when those lockups end and we see whether insiders want out or are willing to stay put.”
“The ones that hold up will start buying. If these stocks stay stable after lockup, expect them to turn into consolidators, using their equity and cash to acquire smaller players. The fintech landscape could look very different by the end of 2026.
“We are absolutely in an AI bubble, but it is uneven. This is a “haves and have-nots” market. Roughly 90 percent of AI companies are probably overvalued and the rest undervalued. The hard part is figuring out which is which. Not all AI bets are created equal.”
“Foundation models look fine. The real bubble risk is in the application layer. The big platforms like Anthropic, OpenAI, and DeepMind are likely fairly valued or even undervalued and should keep gaining share. The crowded part of the market is the thin “AI for X” apps built on top of them, where a dozen companies may be chasing the same niche with very little defensibility.”
Jeremy Kranz, Founder and Managing Partner, Sentinel Global
“It’s hard to be definitive about AI holding up the entire economy, but what we call the circular economy, which is AI companies and data centre companies investing in each other, has a trickle-down economic impact. That would be all the general contractors, all the new housing units for all the people running these data centers, all the people that have to be hired in that area, and all the restaurants that have more business, and so on. So, when you’re talking about potentially $1 trillion of spend happening in the economy around one particular theme and sector, and that is AI, recognising the trickle-down economics does support the belief that, in fact, we have the entire US economy propped up.”
“So, is the economy ‘up’? That’s debatable. What we do know is that the stock market is up and AI is responsible for a huge portion of that momentum, at a scale that’s almost unprecedented. It resembles a modern gold rush where the entire market is tilted toward one sector.
“Is that justifiable? Yes. Will it generate real trickle-down effects? Also yes. But when that flywheel eventually slows, we may discover that the underlying economic realities haven’t moved in lockstep. There’s a broader rebalancing underway, and it’s unlikely we simply revert to any previous economic baseline ever again.”
