Artificial intelligence continues to dominate headlines and fill up investor portfolios, but is the AI hype sustainable or speculative? A new take paints an unsurprisingly nuanced picture: AI isn’t a binary boom or bust – it’s a bit of both.

“We’re seeing record capital inflows, sky-high valuations, one-sided sentiment, and investing driven by FOMO before common sense. Yet we’re also seeing real-world use cases for AI and infrastructure investment at an industrial scale,” says Dan Buckley, Chief Analyst at DayTrading.com. “The best framing is generally that AI is a real boom containing localised bubbles, not a mania across the board.”

A bubble is when the price of an asset, such as a stock, or in this case an entire sector, climbs far beyond its intrinsic value, largely due to investor exuberance and herd mentality in place of raw fundamentals like earnings or credible demand. So, are we in an AI bubble…?

1. Are prices detached from fundamentals? Answer is ‘Yes’. Many of today’s top AI companies are trading at aggressive price-to-earnings and price-to-sales multiples way above historical norms. Of course, some are justified by strong profits, hats off to Nvidia and Microsoft, but the valuations of many newer entrants reflect future earnings that may not materialise. A whopping $560 billion has been invested into AI by major firms in the past two years, yet the estimated incremental revenue is just $35 billion. That’s a monumental gap.

2. Is market optimism outpacing reality? Answer is ‘Yes’. Many corners of governments, corporates, and even wider society now assume that AI will transform virtually every industry. But earnings are yet to catch up. Investors are pricing exponential returns into technologies that are still very much in the early adoption phase. We’re also increasingly seeing companies with limited AI capacity adding AI lingo to their products and marketing efforts to boost their value, a tactic known as ‘AI washing’.

3. Is excessive leverage being used? Answer is ‘Yes’, but it’s uneven. Established players like Nvidia and AWS from Amazon generally fund growth through strong cash flows. However, many AI startups like CoreWeave and xAI are backed by venture capital or debt financing. This is high-risk and could turn sour if interest rates rise or capital dries up. Companies might then struggle to refinance, meet their debt obligations, or even maintain their operations.

4. Is the sentiment one-sided? Answer is ‘Yes’. Bullish sentiment towards AI is dominating both institutional and retail markets. Rarely are we seeing bearish voices being taken seriously, and this creates the potential for sharp corrections if confidence breaks. While Bank of America notes that bubbles often come with rising volatility, the S&P 500’s volatility, so far at least, has remained relatively low. This points to surface-level stability, though it could also reflect complacency.

5. Are inexperienced buyers flooding in? Answer is ‘Yes’. Retail activity in AI stocks is surging. Many investors are buying based on news headlines, social media buzz, or through thematic ETFs. Worryingly, this is in place of running their analysis and studying the underlying earnings. This is much like the behaviour seen in the dot-com bubble, when new money tended to follow compelling storylines rather than raw fundamentals.

6. Are we in a liquidity-driven environment? ‘Somewhat'. Interest rates have indeed risen from pandemic-era lows, yet there is ample liquidity in capital markets. Large tech firms like Microsoft and Nvidia are still enjoying access to cheap financing and piles and piles of cash, meaning they can invest in AI infrastructure without needing fresh equity or to pursue risky borrowing.

7. Is there speculative hoarding behaviour? Answer is ‘Yes’. Companies like OpenAI and CoreWeave are hoarding AI chips, computing capacity, and even engineering talent in anticipation of demand that hasn’t fully materialised yet. Most concerning, in some cases, this spending is taking place before clear business models or an ROI is established. This creates a serious risk to capital if demand growth hits the brakes.

Despite some red flags, the findings make clear that AI comes with some noticeable differences from the dot-com bubble. It’s already delivering tangible productivity gains:

• “AI is already being used at scale”, especially in sectors like media, biotechnology, finance, and logistics, contrasting with the dot-com era when usage often lagged behind valuations.

• Some AI companies are generating value today (even if not yet to the level markets are pricing in). Just look at Nvidia and Microsoft – both already profitable and cash-flow positive - unlike many dot-com era leaders that were pre-revenue or deeply unprofitable.

• Technology giants have committed over $340 billion in long-term, structural investments, primarily in chips, servers, and data centres, that should underpin future growth – not just temporary capital expenditure bursts.

• Former Google CEO Eric Schmidt describes “AI as infrastructure for a new industrial era, not just a passing tech fad.” Many of the world’s brightest minds are sold on its potential.

“The mistake is assuming this is just hype. It’s not. AI is real and valuable,” says Buckley. “But it’s when market sentiment outpaces real business results that I begin to worry about the gap becoming dangerous for investors.”

The analysis finishes with key considerations for investors, notably:

• Focus on profitability: Think carefully about investing in firms that rely on debt or hype without clear earnings. You’re stacking risk on an already high-risk investment.

• Diversify across the AI stack: Semiconductors, cloud infrastructure, and data centres arguably offer more defensible long-term value than pure speculative software plays – even if companies like Perplexity are hot on everyone’s lips.

• Limit exposure to over-concentrated mega caps: The top 10 US stocks now make up nearly 40 per cent of the S&P 500, with most of the AI excitement driven by a handful of key names like Nvidia, Microsoft, Apple, and Alphabet. That kind of heavy concentration makes the market more vulnerable if, or should we say when, any of those companies stumble.