As AI continues to advance at pace, the ways that software and data-owning businesses engage with and harness that technology continue to evolve. Against this backdrop, concerns have centred on the potential for AI to disrupt established digital business models. As investors in a number of what we consider to be world-class digital companies such as Experian, RELX, Sage and London Stock Exchange Group (LSEG), investors ask how vulnerable they are to these risks.
Technology shifts throughout history have almost always resulted in casualties, whether that be individual companies and, in some cases, entire sub-sets of industries. This time will be no different, with many digital businesses potentially in the firing line. Data and software businesses that do not own anything proprietary, and which do not perform mission-critical tasks, may struggle to differentiate themselves.
However, contrary to the indiscriminate sell-off across the Information Services sector experienced earlier this year, the picture is more nuanced than simply a blanket "digital Armageddon" and instead draw distinctions between different types of software and data businesses. Some will likely be disrupted by AI, but others will actively benefit.
So, what are the characteristics investors should seek to know which digital businesses could thrive in an AI world?
When it comes to data owners, not all data is created equally – caches of historic, static data seem to us to be more vulnerable to obsolescence, whereas unique, constantly replenishing, live business data at scale is truly valuable. And it is this type of data which is right at the heart of the earnings power of the companies we own in this space.
For example, LSEG, Experian and RELX all possess vast, proprietary datasets that are becoming increasingly valuable in the age of AI. LSEG owns tick-history data covering more than 100 million financial instruments stretching back 30 years, with this extensive dataset continuing to update around 15 million times every second.
Meanwhile, Experian has built a unique database by persuading more than 200 million individuals worldwide to share their personal financial data in exchange for improvements to their credit scores and access to more competitively priced credit. This database is updated a billion times each month.
RELX's Risk division, meanwhile, effectively already operates as an AI business, with most interactions taking place machine-to-machine and its dataset expanding by 400 million interactions every day. AI also appears to be accelerating growth across RELX's other divisions, with growth in its Legal division rising to 10% in 2025 from 9% the previous year, while its Scientific, Technical and Medical (STM) division saw growth increase from 5% to 6%.
Much of the focus and concern has been on AI's ability to replicate and even surpass the output of incumbent enterprise software providers. The argument goes: if the LLMs are able to fill out a tax return in a matter of seconds at a cheap cost, why will small and medium-sized businesses continue to retain accounting software companies like Sage?
It is easy enough for AI to replicate the code underlying Sage's software – indeed, humans have been doing this for many years, long before AI became a mainstream topic in 2023. This is evidenced by the range of low-cost or even free Sage alternatives on offer in the market.
Yet none of these challengers have ever gained any real traction, and this demonstrates that the ‘moat' around Sage's business is not and has never been its code. The company operates within the accounting industry, where trust, reliability and accuracy is of paramount importance. When dealing with complex information, legal obligations and government bodies, 90% of even 99% accuracy is not enough. In the words of CEO Steve Hare:
"In accounting, payroll, tax and compliance: ‘nearly right' isn't right enough. Customers need outcomes they can trust: accurate, auditable, explainable, and under human accountability."
Perhaps the fact that Sage runs the payroll for 50% of all private-sector workers in the UK illustrates how this dependability and trust has been instrumental in creating Sage's deep moat. And when this trust and reliability is combined with the unique, constantly refreshing data Sage holds on millions of small and mid-sized customers (13 billion invoices pass through its platform every year), AI can become a tool for it to derive even more value from the insights and knowledge that it alone has. Far from being an existential threat to the business, AI offers Sage the potential to develop more functionality and use cases from its data set, deepening its relationship with existing customers and opening up opportunities to attract new ones.
This is not a pipe dream. It is already happening. Sage recently reported that its own AI tool has provided 75 million insights to its customers this year, and its Accounts Payable Agent has saved its customers 5 million hours of work in total. Its overall revenue growth accelerated to 11%, with 102% renewal rates, indicating that real growth is already coming from using AI to better serve its customers. Investors are finally beginning to take notice, it seems. As of the end of October, Sage's share price has rebounded more than 40% since the lows during the summer, and there's been a material recovery in the prices of some of our other digital businesses including RELX, Experian and Rightmove.
Of course, there is still much to do. Nothing remains static and all of these data owners and software providers must continue to evolve and innovate to not only remain relevant but thrive in the face of increasingly rapid technology change. But their ownership of differentiated, privately held data assets not available to the likes of Anthropic, combined with their long and proven track records of being able to innovate and harness the power of previous technology shifts, could well result in them being viewed as AI investment winners once the dust has settled.
Click here to find out more: https://www.finsburygt.com/
