Stay informed with free updates
Simply sign up to the Technology sector myFT Digest — delivered directly to your inbox.
Anyone in doubt about the widening impact of AI on US companies should try taking my commute.
The subway line that stops near the FT’s New York office is plastered with adverts from a trading software group warning that it is “not smart” to use “the same AI for pizza recipes and your portfolio”.
At my suburban commuter train station, the pitch is more obscure. “Get through a VDR faster than the conductor can say Ardsley-on-Hudson,” a billboard promises, adding in smaller type: “AI for M&A.”
A web search made clear that the campaign is promoting software that combs through pre-merger documents in a virtual data room. But I find it hard to believe that many travellers would care enough to check.
Colleagues who attended the World Economic Forum in Davos report that the scale of AI-related advertising there was, if anything, even more overwhelming. Technology providers were lining up to impress the corporate and policymaking elite with their progress in harnessing the world’s hottest innovation.
The drivers of this deluge of marketing are clear. Big companies are under pressure to show results from AI after several years of corporate happy talk that has been long on enthusiasm but short on practical details. Despite widespread rollouts of large language models and thousands of pilot projects, 95 per cent of enterprises could show no positive financial impact last year, according to an MIT study.
Most companies are still struggling to move beyond the basics, such as using large language models to help write marketing materials and employing AI customer service chatbots.
Yet enthusiasts, including many chief executives, have been promising that 2026 will be the year that the technology “grows up” and proves its worth. In a recent Boston Consulting Group survey of global CEOs, 90 per cent of them expected to see a measurable return on their AI investment in 2026 and 62 per cent were confident that the technology would pay off.
Tellingly, the surveyed companies planned to double AI spending as a share of revenue and half of the CEOs told BCG that they believed that their job stability rested on getting AI right.
That creates a very fertile market for AI groups, which can claim a specialist ability, whether it is automating an asset manager’s trade reporting or speeding drug development. Those that can trade on a known expertise are particularly attractive at a time when companies are still casting about for ways to harness the technology to spur productivity gains or bring in new revenue.
The London Stock Exchange Group, for example, is offering clients an AI-based product that lets them design and run complicated financial analysis that once would have required much more in-house computing and tech support.
“It democratises the ability to do sophisticated analytics and modelling. You can be a small hedge fund and do some very cool things. You don’t have to be a giant institution to get access,” explains CEO David Schwimmer. If the tools catch on and work, client hedge funds will see trading gains, and LSEG will prosper not only from selling the tools but also the additional data needed to plug into them.
But the window of opportunity could be a narrow one. Corporate customers are shopping around now, but once they find an AI application that they like, it will be that much harder to shift them to a rival service. Competition is rising as the biggest AI groups and existing software giants, such as Salesforce and SAP, race to create their own specialty offerings. Anthropic sent shares in LSEG and other data firms tumbling on Tuesday by launching a suite of AI productivity tools that help automate legal work.
Hence the massive marketing campaigns by start-ups trying to land a few choice customers before the funding starts to run out.
There is much talk about an AI bubble and how and when it might pop. So far the worry has been focused on the companies that are pouring hundreds of billions of dollars into the infrastructure needed to support the technology. Meta shares took a brutal dive last autumn, and Microsoft’s shares dropped more than 10 per cent last week after it announced that data centre spending had surged 66 per cent.
Microsoft CEO Satya Nadella has warned that the tech infrastructure spending boom risks becoming a bubble unless the benefits from AI usage spread beyond the usual Big Tech suspects. My commute suggests that all manner of companies are trying to make that happen, with what feels like increasing desperation. If they fail, the fallout could be far wider than anyone expects.
Follow Brooke Masters with myFT and on X