Will the artificial intelligence hype-cycle end?
Since the advent of the car, railroads or electricity, few technologies have captured investor imagination quite like generative artificial intelligence (GenAI).
The debut of AI tools like ChatGPT has attracted billions in research and development (R&D) and related construction investment at the same time that share market investors have poured trillions into AI-related stocks, inflating valuations to levels that some suggest should make even the dot-com era blush.
NVIDIA, for instance, has seen its market cap skyrocket on the promise of AI hardware dominance, while startups and Big Tech alike command premiums based on vague assurances of future dominance.
For almost two decades, I have explained that investing in new technologies, especially those that claim to change the course of history, is fraught with danger. Why? Because the progress is typically fast, and picking the winner is nearly impossible. Additionally, as is almost always the case, investors buy everything exposed to the particular theme, pushing up valuations of all participants and essentially betting every company will win, even though that is impossible.
According to my back-of-the-envelope calculations, the collective market value of AI-related companies in August 2025 is approximately US$21 trillion to US$23 trillion, based on major public firms’ market caps (~$18.8 trillion for the nine leaders, plus ~$1.2 trillion – $2.2 trillion for the others) and a speculative private company contribution of $1 trillion – $2 trillion.
If we assume that a 10 per cent after-tax annual return is reasonable for such an investment, then these companies need to generate an annual profit of US$2.2 trillion. And that is a long way from happening because the collective net income of the nine leading AI-related companies (NVIDIA, Microsoft, Apple, Alphabet, Meta Platforms, Tesla, Oracle, Palantir, IBM) for the trailing twelve months is approximately US$420.25 billion. And relatively little of that has come from AI-related revenue streams.
It’s on that point that a sobering new report from MIT’s NANDA initiative, entitled, “The GenAI Divide: State of AI in Business 2025,” throws cold water on AI enthusiasm.
According to the report, 95 per cent of U.S. companies are sinking US$30–40 billion into GenAI initiatives and yielding “almost nothing” in return.
So, it’s gotta be time to ask: Do these paltry results justify the massive investments and exorbitant stock valuations? Or are we witnessing the inflation of a bubble that will inevitably burst?
The MIT report draws from more than 150 executive interviews, surveys of hundreds of employees, and analyses of 300 public deployments. It paints a picture of widespread experimentation but minimal payoff. Adoption is indeed rampant – nearly 90 per cent of workers are dabbling in personal AI tools like ChatGPT for productivity boosts.
Yet, when it comes to enterprise-scale transformation, the story crumbles. Only 5 per cent of pilots scale to production or drive meaningful revenue growth, leaving the vast majority mired in what the report refers to as “pilot purgatory.” MIT suggests this is more than a a tech glitch; it’s a systemic failure in integration, where AI tools lack the memory, adaptability, and workflow fit needed to stick.
At the heart of the issue is what some call the “+AI trap” – the misguided approach of slapping AI features onto existing systems without rethinking organisational design. Generic tools thrive for individuals because they’re flexible and iterative, allowing users to refine outputs in real-time. But enterprise versions? They often flop, as evidenced by a corporate lawyer’s anecdote in the report: A US$50,000 bespoke contract analysis tool delivered rigid, subpar results compared to freewheeling ChatGPT sessions.
Employees, already hooked on “shadow AI” (unsanctioned personal use), abandon clunky official systems that can’t retain context or learn from feedback. The result? Billions funnelled into demos and wrappers that executives dismiss as “science projects,” with no tangible profit and loss (P&L) impact.
The report then goes into sector breakdowns and in so doing reveals even starker disparities.
GenAI is reshaping just two industries – Technology and Media & Telecom – where it’s driving structural shifts, cost reconfiguration, and altered customer behaviours. In these two sectors over 80 per cent of corporate leaders foresee hiring cuts within two years, targeting outsourced roles like customer support and admin processing (impacting 5-20 per cent of such work through attrition rather than mass layoffs).
But in the other seven sectors, which includes Healthcare, Finance, Energy and Retail, the tech is “inconsequential.” Pilots abound, but fundamentals remain unchanged, as one manufacturing COO quipped: “The hype on LinkedIn says everything has changed, but in our operations, nothing fundamental has shifted.”
And, so, the investment landscape feels disconnected from reality. Investors have bid up AI stocks on the promise of ubiquitous disruption, yet MIT’s data shows transformation confined to niches. Sure, some startups have reportedly jumped from zero to US$20 million in revenue in a short time by focusing on single pain points – but the report labels these as ‘outliers’. For most enterprises, a “learning gap” reportedly persists: Tools don’t evolve, users disengage, and budgets evaporate without returns.
It’s a bit hard to keep the hype alive when less ‘hyped’ feedback enters the information loop. Low returns simply don’t justify the hype – not yet anyway.
The gap between investment and return suggests over-optimism, where valuations reflect speculative fervour rather than more sober economics and expectations. The current batch of AI hero stocks, however, trade at multiples that assume exponential growth across all sectors, but the report exposes a yawning divide: High curiosity, low conversion.
If GenAI remains stuck in the 95 per cent failure zone, disillusionment will cascade. We’ve all seen it before – tech bubbles pop when promises are unmet, triggering sell-offs as investor sentiment switches from accepting evangelism to demanding evidence.
That said, there might be some hope. According to MIT’s report, the successful five per cent offer a blueprint. But a lot more companies will need to catch on, fast.
Ultimately, I believe, if MIT’s report doesn’t prick the AI bubble, it will serve as another nail in the bubble’s coffin. The report should be a wake up call to investors highlighting the gap between hype and results. Of course, stock market investors can take a while to wake up!
At this stage of the hype cycle, valuations are generous because the narrative is intoxicating: AI as the next internet, reshaping everything while enhancing human productivity. But with 95 per cent of initiatives floundering, it’s a narrative on thin ice. We’ve been here before.
Roger Montgomery is the Founder and Chairman of Montgomery Investment Management. Roger has over three decades of experience in funds management and related activities, including equities analysis, equity and derivatives strategy, trading and stockbroking. Prior to establishing Montgomery, Roger held positions at Ord Minnett Jardine Fleming, BT (Australia) Limited and Merrill Lynch.
He is also author of best-selling investment guide-book for the stock market, Value.able – how to value the best stocks and buy them for less than they are worth.
Roger appears regularly on television and radio, and in the press, including ABC radio and TV, The Australian and Ausbiz. View upcoming media appearances.
This post was contributed by a representative of Montgomery Investment Management Pty Limited (AFSL No. 354564). The principal purpose of this post is to provide factual information and not provide financial product advice. Additionally, the information provided is not intended to provide any recommendation or opinion about any financial product. Any commentary and statements of opinion however may contain general advice only that is prepared without taking into account your personal objectives, financial circumstances or needs. Because of this, before acting on any of the information provided, you should always consider its appropriateness in light of your personal objectives, financial circumstances and needs and should consider seeking independent advice from a financial advisor if necessary before making any decisions. This post specifically excludes personal advice.
