The world of Artificial Intelligence (AI) is changing daily, bringing with it both excitement and noise. For accounting and advisory professionals, navigating this new landscape can feel overwhelming. In a recent discussion, Richard Francis, CEO of Spotlight Reporting, and Inbal Rodnay, AI Keynote Speaker, explored how accounting firms can move past the hype and into a practical, productive relationship with AI.

Beyond the Hype: A Dose of Reality

It's easy to scroll through LinkedIn and feel like you're the only one not fully leveraging AI. The reality, however, is that nobody has a fully automated practice run 100% by AI agents and employees. The technology is still in its early stages and while AI can perform tasks like reconciling transactions or drafting advice, the applications aren't yet mature enough for widespread, hands-off use across entire firms.

Software vendors are actively working to integrate these capabilities into reliable products, minimising what's known as hallucinations—where the AI generates false or incorrect information. As a user, it’s crucial to understand that AI is a predictive algorithm, not a thinking entity, so it can and will make things up if it lacks sufficient data. For example, it might invent tax clauses or financial figures, which is why human oversight and verification are non-negotiable.

‍

The Path to AI Adoption: A Three-Step Strategy

Rather than getting lost in the noise, firms should focus on a clear, actionable strategy. The goal for 2025 and beyond, isn't to be a tech innovator but to become an AI-literate workforce.

Here are three steps to get there:

1. Invest in a secure AI tool: Move beyond informal, ad-hoc use of free tools. Provide a paid, secure business account for every single staff member—including offshore teams and admin staff. The primary choices are straightforward: ChatGPT or the AI platform integrated with your office suite, like Microsoft Copilot or Google Gemini.
2. Establish a clear policy: A simple, well-defined AI policy gives your staff a safe environment to experiment. This policy should cover responsible use, data security, and the need to verify AI-generated content. You can’t expect employees—many of whom are already using AI in their personal lives and at university—to avoid it at work.
3. Prioritise training: The strategy is to upskill your team. Train them not only on how to use the tools effectively but also on how to use them responsibly. The workforce that didn’t grow up with AI needs to learn how to leverage its power while critically assessing its output.
   ‍

From Doers to Reviewers

AI is not just about cutting minutes from a task; it's about elevating the quality of work. Think of AI as an assistant, coach, and mentor for every team member. A new graduate can use it to draft a strategy, get feedback, and improve their work before showing it to a supervisor. The value shifts from doing the work to driving the technology to produce great work.

The fear that AI will stifle individual genius is understandable, but it's a shift, not an elimination. Just as spreadsheets replaced manual calculations without destroying our analytical skills, AI will change our roles. We are all moving from "doers" to "reviewers." so the key is to be informed and engage with the technology now to figure out how to integrate it positively.

‍

The Two Faces of AI: Automation and Agents

AI manifests in two primary ways: workflow automation and AI agents.

• Workflow Automation with AI: This is a low-risk, high-value entry point. You can take existing, rigid automations (e.g. "if this email has a certain word, apply this category") and replace the manual decision-making with AI. This makes automations smarter and more flexible. For example, an automation could use AI to read a job application, summarise key skills, and put them into a spreadsheet.
• AI Agents: This is a more advanced, high-risk application. An AI agent is essentially an autonomous assistant that has been given a set of instructions, tools, and some autonomy to perform tasks. For example, you could create an agent to handle entity establishment. The agent would read client documents and emails, find all the necessary information, and then ask the accountant only for the missing details before sending a perfectly organized request to the admin team.

While AI agents are making a lot of noise, they should be approached with caution. Start by experimenting with them on low-risk, internal tasks, like categorising emails. This helps your team develop the literacy to ask the right questions and assess the security of vendor-built agents down the line.

Ultimately, the human element remains indispensable. Advisory is, at its core, a human-contact sport—one that technology can enhance but never replace. Tools and platforms can streamline processes, surface insights, and reduce the burden of research, but they are only as effective as the professionals who wield them. By staying informed and sharpening your judgment, you position yourself to harness these innovations strategically, strengthening your capabilities while ensuring they remain a complement rather than a substitute. In doing so, you preserve the trust, empathy, and nuanced understanding that define true advisory excellence, while leveraging technology to amplify—not diminish—the value of human expertise.

If you would like to hear more from Inbal and Richard, watch the full recording here.

‍