AI has been a central focus within the finance industry for the past 12 months. As finance teams have experimented with the technology, valuable use cases have come to light – from time savings through to the automation of workflows to making the sector more attractive to graduates.
This has been occurring against a backdrop of economic uncertainty and heightened costs for businesses. With the National Insurance rise and ongoing inflation pushing prices higher margins are squeezed even tighter than before. Bruce Martin, CEO of Tax Systems, reflects that “in this economic environment, businesses need a sharper view of what genuinely creates value. After years of layering new policies and processes onto existing structures, many organisations are left with a knot of complexity that’s expensive to run and difficult to change. In this case, a reset is good business hygiene, and the most effective leaders will take a more disciplined look at which processes are essential and which can be removed altogether.”
He recognises the importance of AI in tackling these issues in the year ahead: “AI will be pivotal in this regard, enabling organisations to do more with less as budgets tighten. Importantly, however, it isn’t all about speed. The real opportunity with modern technology, particularly AI, is the ability to do things that simply weren’t possible before, unlocking new insights, new ways of working, and fresh avenues for innovation.”
The next innovation wave
2025 was the year of experimentation and laying the foundations to enable finance teams to truly embrace AI. As we head into 2026, they will be ready to make the jump and truly start innovating. When it comes to new use cases, Russell Gammon, Chief Innovation Officer at Tax Systems, predicts that “Agentic AI will be one of the most significant shifts over the next year.”
Agentic AI marks a step forward from generative AI, which has been used by the majority of finance teams in 2025. Rather than responding to prompts, Agentic AI works autonomously, with the ability to reason, plan, and adapt to changing situations.
Gammon recognises that trust needs to be built with AI agents, especially in risk-adverse sectors like finance and tax, but explains that “in the early stages, agents can be deployed in low-risk use cases where trust can be built gradually. Data is one area in particular where agents can add immediate value, freeing up significant amounts of time spent on input and number crunching without disrupting existing controls. As confidence grows, those same agents will start to take on more tasks and responsibilities. One area with significant potential is the review process. Agentic AI can support first and second-level reviews, flagging anomalies, highlighting risks and preparing summaries, which will save a significant amount of time for senior leaders, whose capacity is both limited and extremely valuable.”
Making it happen
Innovation is just one part of the puzzle. How can we reap the benefits of the latest AI developments, if we can’t successfully utilise or implement them? Aqilla’s Founder and CEO, Hugh Scantlebury, advocates for not just successful implementation, but taking “a responsible, transparent approach with clear oversight and guardrails. That way, as the year progresses and finance teams rely more heavily on AI outputs, they can be confident that they understand how those outputs are generated. Without that clarity, they risk creating systems that are fast but untrustworthy, with all the accompanying reputational and ethical risks.
“As developers of AI-enabled finance software, we have a responsibility to support users in this shift by ensuring our systems behave predictably, transparently and in ways that reinforce good financial practice,” he continues. “What we don’t want is for 2026 to become the year of over-automation – because customers aren’t asking for fully machine-led accounting. They want clarity, confidence, and tools that remove repetitive work while keeping judgement firmly in human hands.”
The job isn’t over once implementation is complete, however. Too many businesses in 2025 have made the mistake of plugging AI in and expecting it to work perfectly. But teams must be equipped with the skills needed to effectively use it for any AI project to be a success. Charis Thomas, Chief Product Officer at Aqilla, explains how prompt literacy skills and training will be essential in 2026: “Remember how we once had to learn to search the internet effectively by refining queries, judging sources and understanding how information was surfaced? Finance teams will now need to learn how to interact with AI in the same way. This prompt literacy isn’t about tricks or shortcuts; it’s the modern equivalent of learning how to research properly. As with all computing scenarios, the quality of the question shapes the quality of the answer. Or more colloquially, put rubbish in, and you’ll get rubbish out. Developing that skill across finance teams will be every bit as important as the technology itself.”
Yet, all of this effort is wasted if organisations don’t have the correct infrastructure in place. Terry Storrar, MD at Leaseweb UK, explains that to ensure a positive experience with AI technology, “uptime, latency, reliability, security and scale” are essential.
“Data changes in this industry in milliseconds and they are constantly looking at ways to improve performance and security for their clients,” he continues. “Financial businesses need to prove real productivity advantages to provide a competitive edge as this industry looks for new ways to stand out and doesn’t generally just follow trends for the sake of it.”
Keeping security in the spotlight
It is easy to get caught up in the excitement of the latest AI capabilities, but security must always remain top of the agenda for financial organisations. The fact of the matter is that, whilst we’re using AI tools to enhance and simplify our workloads, cybercriminals are doing the same.
Bertijn Eldering, Associate Sales Engineer at HackerOne, explains that as financial institutions “embed AI into fraud operations, underwriting, and customer service, attackers are using the same tools to industrialise phishing, deepfake scams, and vulnerability discovery.
“In 2026, AI-driven cyber activity will scale on both sides of the fence,” he continues. “Financial institutions must move forward with a return on mitigation-focused mindset. Those that fail to put offensive security testing of their AI and third-party attack surface on an equal footing with product innovation will increasingly find themselves on the wrong side of the next headline.”
From momentum to mastery
2026 will mark a clear inflection point, where AI’s role in finance is no longer experimental, but embedded, operational, and delivering value. The focus is shifting from adoption to execution, and ensuring AI is effectively governed and aligned with business strategy will be crucial to success. As systems become more advanced, innovation must be aligned with cyber resilience, ethical guardrails, and trust. In 2026, competitive advantage won’t come from using AI, but from using it responsibly, securely, and at scale.