The age of AI disruption
Riding the AI wave without getting swept away
Looking back to the early days of the internet, it was clear that the technology was completely transformative; however, many investors who simply bought “anything internet” in the late 1990s ended up disappointed. Those, on the other hand, who identified companies with durable business models, and had the courage to stay invested through the noise, did extraordinarily well.
Today, artificial intelligence (AI) feels similar. The underlying technology is genuinely revolutionary, productivity gains are already visible across a range of industries, and the pace of adoption is accelerating.
Indeed, AI is clearly the most significant investment theme of our generation and sitting it out entirely would be a mistake; yet not every company wearing the AI label will be a winner.
The question is thus not whether to invest in AI, but how.
The infrastructure layer: still room to run
Companies, such as Nvidia, designing the specialized chips that power AI systems continue to sit at the heart of this infrastructure buildout, and for good reason. Virtually every major firm across the world is racing to deploy AI, and they all need the same basic ingredients: computing power, data storage and transmission, and the software to run these new applications. Demand for this infrastructure shows no sign of slowing, with cloud spending by the likes of Google, Amazon and Microsoft still accelerating, driven by growing AI workloads.
Exhibit 1: Big tech is betting big on infrastructure
Source: Bloomberg, AllianzGI, June 2026
Exhibit 2: Magnificient seven and China’s terrific ten
Source: Bloomberg, AllianzGI, June 2026
Alongside this, the key Chinese players, sometimes referred to as the “Terrific Ten” are also spending heavily, catching up with the leading US firms and beginning to rival them in terms of valuation growth.
The dominant chip producers have enjoyed incredible earnings growth in recent years, and we believe there is yet meaningful room for further appreciation as AI adoption broadens from early tech adopters to the wider economy. And the same is true for other infrastructure providers – power providers and distributors, data centre operators, and more specialist semiconductor manufacturers – that are continuing to benefit from waves of investment that are still in their early innings.
A more complex proposition: the application layer
Further up the value chain, the picture is less clear. With the companies that use AI to deliver products and services to consumers, investors should certainly still engage, but there is more reason to be selective. Take the online travel industry as an example. AI will certainly make trip planning easier and faster, which on the surface sounds like good news for booking platforms. However, it may well also allow travellers to bypass intermediaries altogether, as airlines, hotels, and other providers deploy similar technology. The winners such scenarios are less clear and will be harder to call.
Across many industries – such as travel, but also media, professional services, financial advice, heath care, and others – AI will undoubtedly create genuine value. However, identifying which companies will capture that value and convert it into sustainable profit growth presents a more difficult question. Here, enthusiasm for the undoubtedly revolutionary nature of this new technology will not be enough; careful analyses of competitive positions will be the order of the day.
Three questions to ask about any AI investment
- Can this firm charge more, not less, because of AI?
The real AI winners will use the technology to become more valuable to their customers – not just cheaper to operate. If AI makes a company’s product easier to replicate, that’s a warning sign, not a selling point. - Will efficiency gains flow to shareholders?
AI can dramatically cut costs. Yet, in competitive industries, those savings often get passed on to customers through lower prices rather than showing up as higher profits. Seek out companies with enough pricing power to do both. - Is the “data moat” really a moat?
Many companies claim their proprietary data gives them an unassailable AI advantage. We are sceptical. As AI tools improve, the value of raw data is falling and a lasting competitive edge needs to be built on something harder to replicate – a brand, a network, a regulatory position, or a deeply embedded customer relationship.
With opportunity comes risks
While the opportunities around this new technology are clearly significant and not to be ignored, some sectors are facing genuine structural headwinds that can easily be underestimated amid the broader noise and excitement. For instance, a law firm or financial back-office operation that loses a meaningful share of its workload to AI tools is likely not facing a temporary problem, but rather a permanent change to its business model. And the same applies across many other sectors, such as certain media operations and other knowledgeintensive businesses. The key here is certainly not to ignore these sectors entirely, but to appreciate that the evidential bar high in terms of ensuring that management and leadership teams have a credible plan to adapt to the changes that are coming. Or, in many cases, are already here.
In addition, we are now currently amid a wave of mega IPOs – including SpaceX, OpenAI, and Anthropic – set to bring some of the world’s most influential private AI companies into public markets. While markets will likely be able to absorb these large listings without significant disruption, their arrival could redirect investor attention away from existing beneficiaries toward newly listed leaders. And although this wave is certainly expanding investment opportunities, high valuations and elevated expectations mean selectivity remains crucial here too.
The bottom line for investors
AI is clearly too important and fast moving to sit on the sidelines. However, the spread between the genuine long-term winners and those companies that are simply riding a wave of enthusiasm is wide, and likely to widen further as the technology continues to mature.
Our approach here is to stay invested in the theme, concentrate on those companies with real earnings momentum and durable competitive advantages. Meanwhile, staying alert to disruption risks, even among today’s most celebrated names, will remain key.
With AI, as with investment around any great technological shift, selectivity is what will separate return from regret.