Discussion around Artificial Intelligence (AI) and commercial property has hitherto often focused on whether AI accelerates employees' return to the office. However, this is the wrong question. It is more important to analyse where AI companies are investing, hiring staff, and expanding, as these decisions reveal more about the next office cycle than any return-to-office policy. After 20 years in commercial property, one learns to be sceptical of bold predictions. Many cycles have been overstated, but this one seems different, and the data supports this.
First, let's consider where demand is actually concentrated. One submarket in San Francisco – the non-central business district, including South of Market and Mission Bay – accounts for 25 per cent of all active AI office demand in the markets tracked by VTS. San Francisco's South Financial District contributes a further 16 per cent, San Jose 14 per cent, and New York City's Midtown South and Midtown 11 per cent and 8 per cent respectively. These five submarkets across three markets together comprise almost three-quarters of current AI office space.
Concentrated Growth and Changing Requirements
National AI office demand has increased by 85 per cent year-on-year, but in the industry's largest AI centres, the increase was 179 per cent. The development is not just growth, but concentrated growth. Today's market is fundamentally different from the last technology cycle. OpenAI raised $122 billion in February at a valuation of $840 billion. Anthropic followed with $30 billion in a Series G at $380 billion. These are not companies from the 2022 era burning through staff against speculative valuations – both are scaling their revenue significantly faster than their headcount.
The current fundamentals are more robust and the market opportunities significant. Jensen Huang, CEO of Nvidia, recently stated that writing code is not the job of a software engineer; they are and always have been problem solvers. This understanding is crucial to comprehending the concentration of AI demand in the Bay Area and New York City. The companies expanding in these markets are not looking for junior programmers, who are available everywhere. Currently, the demand is for experienced problem-definers, i.e., those individuals who decide what is developed and why.
- —AI does not diminish the value of exceptional talent, but amplifies it.
- —Such talent naturally clusters, and AI has made these clusters even more valuable, as a single experienced problem-definer can now direct far greater output than ever before.
- —The next generation of engineers will benefit: AI gives a junior engineer leverage from day one that previously took a decade to acquire.
Blackstone and Brookfield recently invested a combined $2 billion in enterprise AI implementation platforms based on OpenAI and Anthropic. When companies of this magnitude make such structural commitments, it means they have already identified productivity gains and believe in the next phase of large-scale implementation. This implementation will not happen solely in the Bay Area and New York. Blackstone and Brookfield's portfolio companies include energy, finance, utilities, industrial, and healthcare sectors. Integrating AI into these companies creates demand in the markets where these companies are already operating.
Emerging Hotspots for AI Demand
VTS demand data already shows where this next wave is beginning. Seattle is the most visible early signal, with a 390 per cent year-on-year increase in AI demand, driven by companies expanding into an engineering talent market that is deep and more cost-effective than the Bay Area. Austin is another market, now with 1.2 million square feet of active AI demand, making it the fourth largest AI market. Northern Virginia, Atlanta, and Chicago are the markets to watch for the next phase of development. Each of these markets reflects a different driver of AI adoption, from defence and government technologies to enterprise software and broader business solutions, but together they show where demand is heading next.
For commercial property executives, portfolio decisions – where capital is allocated, how assets are positioned, and which markets are prioritised – should be informed by forward-looking indicators rather than lagging market statistics. The demand growth in Seattle, Atlanta, Northern Virginia, and Chicago is already telling a story that market statistics will only confirm in six to twelve months. The more important question for executives, however, is what AI means for their own organisation and whether they are building this capability on the right foundation. Blackstone and Brookfield have made these commitments because AI is already transforming the way their own firms operate. The companies best positioned for this cycle are those whose data is sufficiently structured for AI and who laid this foundation before they needed it. This is what makes AI adoption a sustainable advantage.
The office market has always rewarded those who recognise structural changes before they become obvious in market data, and AI is creating just such a moment. Helping to build this foundation in the industry has been VTS's mission from the outset, and we believe this will become even more important in the age of AI. Ryan Masiello is Co-founder and Chief Strategy Officer of the property software platform VTS.
