AI: from uncertainty to opportunity
Bob Hunt
25 November 2025The question is no longer ‘if’, but ‘how well’, we adopt it. Our recent survey of the broker community gave a clear read on where advisers are today and what they need next. The message is unmistakable: interest is huge, usage is growing, but there’s a real education and implementation gap we must close quickly.
Three numbers tell that story. Firstly, 81.7% of respondents have used AI tools like ChatGPT or Copilot in the last three months. Advisers are hands-on; they’re experimenting and seeing the time savings first-hand.
Secondly, 75% say they understand how AI has evolved over the past 50 years. There’s a base level of knowledge that’s far better than some might assume. But thirdly, only 45.1% understand the difference between open and closed data – an essential concept for anyone handling sensitive client information.
That gap matters. Without robust understanding of how data is sourced, moved and protected, firms risk using powerful tools in ways that are inefficient at best and non-compliant at worst.
We asked firms to rate their progress – using a scoring of one to five – across five pillars, and the results showed that there is clearly work to be done: aligning AI with business goals (2.45), building an actionable roadmap (2.26), optimising the tech landscape (2.58), data governance (2.82), and empowering the workforce (2.47).
These are steady ‘work in progress’ scores, but they show an industry that has just started the journey and one that needs structure, training and governance to scale safely.
Marketing is a good barometer of practical adoption. 29.9% of firms already use AI in marketing, predominantly for content creation and ideation (40.9%), content production (22.7%), design tools (13.6%), and communications/analytics (each 9.1%), with a smaller slice using it for lead generation (4.6%).
That’s a clear signal: wherever there’s repetitive, time-consuming work, AI is gaining traction first. And it makes sense – faster campaign production, better consistency, and the ability to personalise at scale are tangible wins.
But we should not confuse experimentation with readiness. 86.75% of respondents want to learn more about AI and how it can work for them. They specifically ask for help on process efficiency, marketing strategies, lending and verification, customer experience and foundational AI knowledge.
Risk must be a consideration
In other words, advisers aren’t resisting change, they’re asking for the playbook. That’s the education gap we must close: we need to help advisers answer questions around which tools to use, for which tasks, under what controls, and how do we measure value and risk?
Risk also can’t be an afterthought. We are entering a more complex era where the same technology that helps us can be weaponised by bad actors. Synthetic documents, fabricated identities and convincing AI-generated communications are already out there. Controls around document authentication, source verification and auditability need to be designed in, not patched on.
Here’s the practical reality for advisory firms. There’s a world of difference between fearing what AI might mean for the role of the adviser and using it to do the job better.
Firms have an opportunity to: treat AI as a disciplined operating upgrade; automate low-value tasks, standardise and speed up marketing, use summarisation to accelerate case notes and file checks, deploy pattern-spotting tools to highlight anomalies before they become problems, and underpin it all with data governance, training and an internal policy that staff actually understand.
At Paradigm, we’ll help deliver that discipline. We’ll look to issue guidance and information on safe, compliant use of large language models (LLMs), practical marketing use cases, and the fraud prevention basics every firm should adopt.
We’ll also convene lenders and brokers to share what works, where value is being created, and where the red lines need to be. The goal is simple: help advisers turn curiosity into capability, and capability into consistent client outcomes.
The takeaway from our survey is encouraging: advisers are leaning in. They want to learn. They see the upside. Now it’s on us to provide the training, standards and guard rails that let firms scale AI confidently.
Today is the worst AI will ever be. If we build the right foundations now, we won’t just keep up, we’ll set the pace for an advisory model that is faster, safer and demonstrably better for clients.
