The key question over these past two years has been whether or not the massive investments in data centres, chips and computer models would translate into equally impressive returns; the latest earnings season has brought with it the proof.
It is in the field of cloud computing that this is most compelling: the growth of the three main hyperscalers rocketed in the second quarter of 2026, with AI being increasingly named as the biggest driver. Over the first half of the year, Google Cloud’s revenue growth surged from 63% to 82% year-on-year, Microsoft Azure’s (at constant exchange rates) grew from 39% to 43%, and AWS’s (the leading cloud platform) went from 28% to 37%, which is a quite remarkable rate given its annualised revenue USD 169 billion.
This acceleration is not simply circular financing within the tech sector; AI take-up has spread across all industries. Microsoft’s 365 Copilot has over 30 million paid licences, or around 6.5% of Microsoft 365’s commercial licences. Penetration remains modest, but the trend is encouraging: the number of paid Copilot licences has more than doubled since the end of 2025 (up from 15 million to more than 30 million in June 2026). The sequential growth of licences is also picking up pace, going from 30% to more than 50% – a sign that companies are shifting from testing to more widespread roll-outs.
Final demand growth is most evident in model providers. Annualised revenue figures for Anthropic have risen from around USD 9 billion at the end of 2025 to more than USD 65 billion in July 2026, while the latest annualised revenue for OpenAI has already breached the USD 40 billion mark. Doubts remain about the profitability of base models, especially at OpenAI, but such a rise in revenues shows that clients are ready to hand over significant sums for AI capabilities.
This, however, poses the question: is the monetisation of AI moving at sufficient speed to justify the size of these investments? Company directors are optimistic. Alphabet released its investment forecasts for 2026 (USD 195–205 billion) and expects these figures to increase further in 2027. Amazon again upped its 2026 target to USD 220 billion and indicated that its capacity will be limited into 2027. Microsoft is growing its computing capacity at pace, with demand for Azure continuing to outstrip supply. Even SpaceX is becoming a major player in AI infrastructure, with Elon Musk aiming to roll out 15 GW of capacity by the end of 2027.
The debate about monetising AI is far from over. Depreciation, energy costs, and interest charges will continue to rise. The rate of uptake is not linear, and data centre financing remains a challenge. However, the burden of proof is shifting: a year ago investors had to put their faith in demand projections; today, the expansion of the cloud, the growing uptake of AI by businesses, and soaring revenues at the main providers of models provides solid proof that investments in AI will be profitable.
The opinions expressed herein are correct as at 20 August 2026 and are subject to change without notice. This information should not be relied upon by the reader as research or investment advice regarding any particular fund, strategy or security. Past performance is not a guide to current or future results. Any forecast, projection or target, where provided, is indicative only and is not guaranteed in any way.
