This FTSE 100 growth stock was widely regarded as one of the market’s great long-term compounders over the past two decades. A steady business model, strong margins, and consistent earnings growth helped build that reputation.
However, with the shares now down around 36% over the past year, investors are beginning to question whether that growth story is still fully intact.
AI threat to the growth narrative
For years, Relx (LSE: REL) has been viewed as a classic compounder, driven by sticky customers, high margins, and a strong position in data and analytics markets. However, that long-term growth narrative is now facing a new kind of test.
The concern in the market isn’t about short-term performance, but structural disruption. The rise of AI-enabled workflow tools has raised questions about whether large technology companies could begin to replicate or replace parts of the software and information services that have underpinned the business’s growth.
In particular, investors are asking whether generative AI could weaken pricing power, reduce switching costs, or allow customers to bypass specialist platforms altogether.
In other words, if AI can increasingly deliver research, analytics, and workflow functionality in a more automated way, does the traditional moat still hold?
That’s the debate now sitting at the centre of the investment case for Relx.
Durable moat
If AI can now summarise, search, and generate information, what exactly stops customers from switching away from RELX’s offerings?
On the surface, the threat sounds credible. Lawyers are increasingly experimenting with AI-enabled workflow tools such as Harvey and Legora. That naturally raises the question of why firms would pay for multiple overlapping systems when newer tools appear faster or cheaper.
But that framing misses how these platforms actually work in practice. Large law firms don’t operate a single digital ecosystem. They use dozens of different tools depending on jurisdiction, task, and client need. In that environment, AI tools aren’t replacing core legal research platforms — they sit alongside them, layered into existing workflows.
To me, the key point is where the real value sits. It remains anchored in curated, verified, and continuously updated legal and regulatory content.
AI may change how users access information, but it doesn’t easily replicate data ownership, trust frameworks, or the regulatory standards that underpin legal and compliance work.
A similar dynamic exists in risk and scientific information services. There, a large share of revenues comes from machine-to-machine systems built on decades of proprietary datasets and embedded regulatory infrastructure.