FutureProof: What are the professional indemnity implications through accountants’ adoption of generative AI?

In the third article of our FutureProof series, we shall explore how accountants are adopting generative artificial intelligence (Gen AI) and how this might impact on the types of claims we will see and the risk profile of firms.

The accountancy profession is extremely broad

When we talk about the accountancy profession, we need to bear in mind that it is incredibly broad and diverse encompassing radically different work and client profiles, from one of the thousands or so high street practices focussing on tax returns and book-keeping for small businesses to the likes of the international firms such as PWC with £billions of UK revenue specialising in a huge range of corporate, audit and advisory work for multi-national corporations. So, when we think about accountants adopting technology and AI, it’s important to remember that firms will adopt technology in different ways in according to the scale and nature of their business. The professional indemnity outcomes and impacts are therefore likely to vary across the market.

However whatever the type of firm adopting Gen AI, its core intention, at heart, is the same - to automate processes and increase productivity.

To what extent are firms already adopting Gen AI, and how?

Thomson Reuters have done some interesting research this year asking accounting respondents from the UK, the US, Canada and New Zealand about their adoption of Gen AI. Something like 8% said they were already using Gen AI and another 13% said they would do so. 30% were considering it and 49% had no immediate plans. So it was on the firm radar of roughly 50%. Other research published earlier in the summer this year in Accountancy Age indicated an incredible 99% of accountants responding to a survey acknowledged that they had at least used some form of basic AI to provide a service to their clients.

The very large accounting firms, as you would expect with their huge resources, are typically early adopters of technology and Gen AI is no different. They are investing enormous sums into Gen AI tools (which has the feel of a space race) in order to drive profitability and efficiencies. What are they doing differently?

In some respects they are using similar tech to the large law firms such as automated document review platforms to distil key information from a suite of documents. Notably, they already use Gen AI based predictive analytics to help their clients make data-driven decisions; this might be trend forecasting, growth strategy and risk profiling.

Interestingly is the more bespoke use of Gen AI for audit work – in simple terms they have tools to not only extract key data from servers/data banks but also to analyse it, generate insights and support decision making. This uses Large Language Models (LLMs) connecting the objective hard data with an organisation's knowledge base, the subjective. Importantly, they are implementing techniques to ground the source of the data to drive accuracy and avoid data breaches.

Like smaller firms they are also using it for tax planning and advisory work – at the least to do some of the heavy lifting before advisers provide more traditional analysis.

We would expect the practices of the early adopters to continue to flow down the profession with medium firms adopting technology and the techniques for using it. This in turn will hard-wire use of Gen AI in accounting practices for the future.

What does all this mean for the professional indemnity claims landscape?

We have been commenting for a number of years about how change is coming in both the risk profile of accountancy practices and the type and intensity of future claims. Obviously, we should be mindful that the claims landscape for smaller and medium firms is different to that of the very large practices but in our view, change will be more demonstrably felt by small and medium firms, and their participating primary insurers.

For some time, smaller firms have been seen as relatively low risk, with a low premium – this reflects the day to day accounting work they carry out from book-keeping to tax returns. Whilst this might generate a predictable flow of claims across an insurer’s book, they have tended to be very low value. However, automation will likely iron out these attritional claims which typically arise from basic human errors.

So at face value, will the risk profile of these smaller firms become more benign? Possibly in the short-term. However, we suspect the real impact will be felt some years away as the adoption of Gen AI forces a change in the type of work firms practice in addition to how they do the work. How might this play out? Firms will be competing for less work as a result of automation (which will also include clients handing some of the processes in-house). So, firms may well face extinction or expand into new practice lines. Research tells us that principals in those firms know that the adoption of Gen AI is a double edged sword – they first need to adopt or die but as a result they will then need to adapt or die.

In simple terms, we suspect there will be a race upstream to adapt to providing more advisory and complex work handled by the larger firms – the type of work which AI will make more efficient but is unlikely to replace. From a claims perspective, that is the riskier work which leads to higher value and potentially higher volumes of claims. In these circumstances, the risk profile of firms will change as will their premiums. We might see more consolidation across the profession and quite possibly a reduction of the typical small high street firm. Underwriters will also need to adapt to the evolution of both the profession and the risk profiles of firms.

As far as the larger firms are concerned, they in turn will be chasing the more complex work as automation harmonises their processes. That will give rise to errors by accountants practising in areas they are less familiar with. As with the smaller firms, though, we might see fewer claims in certain situations where human error has been ironed out. We might also see more efficient and non-negligent auditing, which more easily spots underlying fraud and more efficiently selects the scope of the review. Audits claims are low in volume but destructive in value. If AI can remove some of the indemnity risks from auditing, that might make certain firms a more attractive risk for underwriters and hopefully reduce premiums. The same might be said for the complex tax advisory work where large claims can arise out of basic failures to give adequate risk advice.

There will be other losers from the adoption of Gen AI, where firms do not implement proper guardrails to protect themselves and importantly their clients, from the careless use of AI giving rise to data and confidentiality breaches as well as negligence claims. We will explore in a future article how claims might evolve against accountants’ adoption of Gen AI.

Change is coming - we just don’t know exactly what that will look like

The reality is that much of the above is conjecture and we do not know how this revolution in working practices will impact firms and their insurers. However, it is a safe bet to contend that the years of a predicable accountants claims landscape are coming to an end but that change won’t be straightforward or linear. On the whole, it's our view that Gen AI will ultimately be a positive thing for the accountants PII market but that does not mean there won’t be a few bumps along the way. Understanding and adaption will be key for both firms and their insurers.

Our content explained