[SPONSORED CONTENT]
Growing AI use is shifting attention to responsible, large-scale adoption.
Over the past few years, AI has become a fixture at financial services conferences. Pilots, innovation projects and bold predictions about its potential to transform investing, saving and pensions have been hard to miss.
But 2026 feels different.
This is the year UK financial services stopped talking about AI as a future trend and started treating it as a present-day reality.
In many ways, it reminds me of the journey we have been on with social media.
Consumers embraced social media long before most financial services firms fully understood how it would influence financial decision making. Eventually, providers, advisers, and regulators adapted, and what started as a new communication channel became a regulated part of the financial services landscape.
A recent FCA case underscored this shift, with seven social media influencers sentenced for unauthorised financial promotions.1 The FCA has also led wider action against so-called ‘finfluencers’, making clear that consumer protections apply regardless of whether financial content appears on a website, in a newspaper or through social media feeds.
AI feels like it is following a remarkably similar path.
Consumers are already using it. Providers are rapidly deploying it. Regulators are responding. The conversation has moved on from ‘should we use AI?’ to ‘how do we use AI safely and responsibly?’
Why 2026 feels different
What has been striking is not any single development, but the sheer volume of activity across the first nine months of the year.
The FCA launched the Mills Review into advanced AI, consumers, markets and regulation. It expanded its AI Lab activities, reopened its AI Input Zone and published a major review of how regulation may evolve in an age of increasingly autonomous AI.2 The shift was reinforced in June by Nikhil Rathi, Chief Executive of the Financial Conduct Authority, in a major speech on regulation in the age of AI.3 Scottish Widows was subsequently included in the second cohort of the FCA’s Anthropic Supercharged Sandbox as the sole pension provider selected to participate.
Meanwhile, The Pensions Regulator published its AI Plan and set out clear expectations for pension schemes, trustees and administrators on governance, oversight, data quality and member protection.4
Taken individually, each announcement was noteworthy. Together, they signal something bigger.
The regulatory focus is no longer whether firms use AI. It is about how they govern it, monitor it, test it and remain accountable for the outcomes it creates. As AI becomes more capable, regulators want evidence that firms understand the risks as well as the opportunities.
For those of us working in workplace pensions, it feels very much like a ‘let’s get serious about AI’ moment.
Consumers aren’t waiting
One of the most interesting findings from the FCA’s work is that consumers are already bringing AI into their financial lives.
Research highlighted during the Mills Review found that nearly a third (29%) of people who had sought pension advice in the previous 12 months had used AI.5
That chimes closely with what we found in the Scottish Widows Retirement Report 2026 when we explored the role of technology and AI in retirement decision-making.
Our research found 30% of people would trust AI tools to provide guidance about their pension. However, perhaps the more important finding was around who they trust.6
Of those willing to trust AI, four in five said they would place most trust in AI that was backed by a pension provider or regulated financial expert. Just one in five placed greater faith in technology companies.
The findings suggest people are open to AI but want safeguards. They told us they could see clear value in AI helping explain pension jargon, estimate retirement needs and support everyday financial decision-making. However, they had concerns around unsuitable recommendations, personal data and whether AI would fully understand their individual circumstances.
What struck me when comparing the Scottish Widows research with the conclusions of the Mills Review was not how different they were, but how similar.
Both point to a future where AI becomes a normal part of financial decision-making. Both highlight trust as the critical ingredient. And both suggest the immediate opportunity is not replacing advice but helping people understand their options before engaging with guidance, advisers or providers.
Learning from real life
In the first half of 2026, Scottish Widows launched the InvestAI agent, which has since supported more than 8,000 conversations with pension scheme members.
For us, AI is far from theoretical: it is helping real people with questions about pensions, investments, retirement and financial wellbeing.
Every conversation teaches us something: how customers naturally talk about pensions, where jargon creates barriers, what topics generate uncertainty – and where people want more support. In many respects, the learning is as valuable as the technology itself.
One of the big learnings from InvestAI is that people often want help understanding rather than deciding. They are looking for explanations, clarification and confidence.
The difference is that, unlike general-purpose AI, this agent operates in a regulated environment, with clear accountability for the answers it gives.
AI arrives in trustee boardrooms
Trustees feel the impact of AI without deploying it directly themselves. They increasingly report seeing AI- generated content in Internal Dispute Resolution procedures, as members use it to draft complaints, challenge decisions and organise detailed submissions.
This can help members articulate their concerns and engage with complex issues that might otherwise be difficult to explain.
But AI can also produce convincing arguments built on misunderstandings, incomplete information or assumptions that do not apply to a particular scheme. It is making AI governance highly relevant to trustee boards, as well as providers and administrators.
What should providers and trustees do now?
Understand where AI is already being used across the organisation and supply chain.
Make accountability clear. AI may support decision-making, but responsibility cannot be outsourced to an algorithm.
Strengthen governance, testing and ongoing monitoring.
Equip administration and customer-facing teams to deal with AI-generated queries and complaints.
Be transparent with members about when AI is being used, what it can do and where human support is available.
The future is closer than we think
The Mills Review talks about a future where AI does more than answer questions. It recommends preparing for a world in which AI systems recommend actions, initiate transactions and operate within agreed permissions on behalf of consumers.
This is often described as agentic AI, and it sounds futuristic.
Yet we have already come across a customer who built their own AI agent and used it to interact with the insurance part of our business in pursuit of a better deal.
This is why I think we underestimate the pace of change. When people talk about the future, they often imagine something decades away. But reality may be much closer than that.
As we move towards the end of the ‘Connected Twenties’ – the decade when digital became everyday – we can already see the building blocks coming together: pensions dashboards, connected finances, trusted digital identities, AI-powered guidance and, eventually, agent-to-agent interactions.
A few years ago, I wrote that pension savers are fundamentally trying to answer three simple questions:
What have I got?
Is it enough?
What can I do next?
Pensions dashboards will make it dramatically easier to answer the first question. AI can help people understand the second. And agentic AI may eventually help people act on the third.
That is an exciting prospect.
But if 2026 has shown us anything so far, it is that the future of AI in financial services will not be determined by technology alone.
It will be shaped by trust, governance and the confidence people have that these powerful new tools are genuinely working in their interest.
Read the latest news, expertise and thought leadership from Scottish Widows’ workplace pensions experts –here.
To read more articles from Scottish Widows, visit our content hub on Corporate Adviser.
____________________________________________________________________
1 FCA, Influencers fined for issuing unauthorised financial promotions, February 2026.
2 FCA, AI and the future of retail financial services (The Mills Review), July 2026.
3 FCA, Rethinking regulation for the age of AI, June 2026.
4 The Pensions Regulator, AI Plan, May 2026.
5 FCA, The Mills Review, AI and the future of retail financial services, page 55. July 2026.
6 Scottish Widows, Scottish Widows Retirement Report 2026, Part Three: technology and AI in retirement decision making – Taking a closer look at AI, August 2026.