Is the AI sector in a bubble? It depends on who you ask. While some figures in the industry have been adamant the fundamentals of the AI industry are sound, it is not difficult to see why fears around a bubble persist.
There are massive valuations, huge funding rounds and a seemingly unending parade of companies claiming to have the best AI tools to help us transform our business and personal lives. In February, OpenAI announced a $110 billion funding round, with Nvidia chipping in $30 billion and Amazon $50 billion. OpenAI’s valuation in October last year was $500 billion; the most recent figures put it at around $730 billion.
Even those at the top seem to be on the fence a little. Google’s Sundar Pichai has described AI as having an “extraordinary moment” but said there were “elements of irrationality” in the market. In an interview last year, OpenAI chief executive Sam Altman acknowledged that investors were “overexcited” about AI and warned someone would lose a “phenomenal amount of money”. But on the other side of things, he said “a lot of people are going to make a phenomenal amount of money”.
We have been here before. The dotcom boom of the late 1990s became a bubble that burst in the early 2000s, taking plenty of promising businesses with it and leaving investors wary.
READ MORE
But from that difficult period came a new wave of innovation, with some of the stalwarts of the tech sector – Google, Amazon, and so on – surviving the dotcom bust with a strategy that propelled them into a dominant position in the market. What followed behind them was a wave of born-on-the-web companies that created their own markets and exploited them for a decade or more.
Could the same thing happen with AI? There have been warning signs about how a market wobble could hit AI companies and their stocks. The release of DeepSeek’s high-performance model caused a major wobble last year in market confidence, leading to a temporary but significant drop in AI stocks. Nvidia stock lost 17 per cent in a single day in late January 2025. Things recovered, but it was a warning shot.
Niall McEvoy, head of venture with Elkstone, compares AI and its transformational potential to the impact on transport of railroads, cars and the combustion engine.
“In terms of what that is, what that did for travel and transport, this is doing for the ability and the power of companies to be able to create technical and commercial advantage at speed,” he says.
The dotcom era had a lot of speculation McEvoy says, with companies moving from the real world online.
“Did anyone really know what they were doing? And did they have the tools to do it at the pace that it was being invested in? Probably not. That definitely was a bubble,” he says.
“What you have here is a rapid inflow of capital into the infrastructure that’s supporting the technology. You have the rapid access to tooling at a basic human level to be able to do things now that, historically, may have taken teams of people years to do.”
The AI sector is not just about chatbots and large language models; there is the hardware sector of the business that has propelled Nvidia, for example, into a $5 trillion market cap.
But there have been concerns about the circular nature of the investments. Companies such as Microsoft, Amazon and Nvidia have been investing billions into the sector. But these same companies will ultimately be their customers, and future valuations will be dependent on realising the potential of profitability.
“There is a little bit of the snake eating its tail,” says EY’s Grit Young. “I think there’s a bit of fomo [fear of missing out] as well. We don’t want to lose and we can’t afford it. Some of the companies can’t afford to, because they will feel that their revenue will be attacked and cannibalised, if it’s not by them, it’s going to be by their competitors. So, there’s also an element of ‘we don’t really have a choice but to develop this’.”
There are few industries that are expected to remain untouched by AI and its potential transformative effects. The breakneck speed of innovation, however, makes it difficult at times to remember that we are still in the early days of this transformation. Competition is rising, which could lead to a sharp correction in the future.
While some companies in the sector are attracting huge valuations that could be vulnerable to a correction, the general thinking is that the AI sector itself will continue to grow. Industry watchers are drawing parallels with the development of other technologies such as cloud computing.
“There is a huge investment in AI at the moment, but it was actually very similar with what happened with the internet, with mobile, with cloud,” says Dorothy Creaven of Irish AI start-up Jentic. “These types of investment cycles always appear during major opportunities or major technological changes. The period that we’re entering into is one where software is going to be built and used very differently.”
However, AI is vulnerable to the weight of expectations. For the past couple of years, companies have been talking up the benefits of AI and how it will revolutionise everything from software development to customer relations. But survey after survey shows a big gap between expectations and reality, with companies hesitant to jump into the AI landscape. A recently published survey from Google found businesses considered a lack of skills, a fear of making mistakes and the costs associated with rolling out the new technology were behind a reluctance to implement AI, even though the majority believed it would benefit their businesses in the long run.
“We get very excited when it’s in our hands, and every consumer can use something like ChatGPT or Claude. But ultimately, from a work perspective, maybe treat it like an intern or a junior employee who can do a lot of these tasks, but the relationship needs to be built up. It’s going to be one of trust,” says Creaven.
In the future, the battle between Europe and the US over AI supremacy could become an obstacle, specifically the power struggle over regulation in the sector.
It is not just about the risk posed to the AI sector. There is also a risk from the AI sector, namely that the current fascination with AI and the fear of missing out will have a negative impact on other industries. Climate tech, and software are two areas that are often cited as seeing a knock-on effect as investor focus shifts elsewhere.
“I think this is very real, and I think it’s definitely going to be sustained, but it is clipping capital available to other sectors,” says McEvoy.
“It’s definitely squeezing more traditional enterprise software plays even, even those that are AI first.”
EY’s Young agrees that some areas that used to attract attention have fallen off the investment league table.
“But that’s probably not just to do with investors wanting to invest in AI, but also other political changes. Using the climate tech sector or sustainability as an example, there’s obviously some political and regulatory changes and incentive changes that have changed investor appetite for that sector, but that’s one where you think, fundamentally there should be a huge amount of demand,” she says.
Regardless of the opinions on different companies, almost everyone agrees on one thing: AI will be transformational for business. As a result, there is a general optimism that the technology will start to pay off in the coming months, for the businesses that are investing in the capabilities at least. In 2025 the talk among businesses may have been around AI strategy; in 2026, that is likely to be what is being delivered by AI and how it is impacting business metrics.
And that may be crucial to the continued growth of the industry. If businesses fail to see enough benefit from the investment in AI, they may reconsider putting further money into it.
For now, however, the investment continues, with companies looking to agentic AI – tools that can carry out tasks independently of human intervention – as the next phase of the industry. That will require more hardware, more cloud services, more data centre space. The cycle continues.
Just don’t mention the B word.
“I think the question about whether we’re in an AI bubble or not ends up distracting us from what’s actually happening,” says Creaven, “which is really a long-term technological shift.”
