In the latest chapter of generative artificial intelligence’s meteoric rise, Mistral AI has just unveiled Europe’s response to OpenAI’s ChatGPT.
Dubbed ‘Mistral Large,’ the French startup unicorn’s new learning language model (LLM) “reaches top-tier reasoning capabilities” rivalling those of its Silicon Valley counterpart according to Mistral’s founders, while the parallel release of the company’s free multilingual conservational assistant ‘Le Chat’ complements its ambition to “put frontier AI in everyone’s hands” – from curious individuals to forward-looking companies.
This transatlantic competition is likely to yield positive developments for AI innovation, laying the foundation on which firms in myriad sectors will be empowered to build new services and future-proof their operating models. While perhaps not immediately associated with AI’s transformative potential, the accounting profession is already beginning to seize the benefits of this emerging tech revolution to create a more innovative, human-focused sector ready to meet the risks and opportunities of an ever-changing economic landscape.
Unprecedented generational shift emerging
As Thomson Reuters’s recent ‘Future of Professionals’ report rightly notes, “navigating uncertainty in the business environment” driven by technological progress has been a reality that “tax and accounting professions have been dealing with, off and on, for decades.” From the widespread workplace adoption of computers in the 1980s to the turn-of-the-century Internet boom, our industry has had to grapple with the fundamental, irreversible changes to the accounting profession triggered by these generation-defining developments.
Although its breakneck evolution has raised more questions than answers, it has become increasingly clear – particularly since the watershed launch of ChatGPT in late 2022 – that the AI era will drastically surpass the scale and impact of previous technological earthquakes. While some accounting professionals are certainly sceptical of this seismic change, most
remain broadly optimistic about AI’s disruption of our field. As the industry’s bullish response to AI reflects, firms that fail to capitalise on AI-driven efficiency, productivity and value-generation gains will be doomed to fall behind in today’s rapidly-evolving climate.
Last May, a Thomson Reuters Institute survey revealed that 15% of accounting firms were already using or planning to imminently adopt AI technologies – a figure which will have grown substantially as over half of respondents reported their AI integration intentions for the ensuing year. Moreover, while professional opinion remains roughly evenly-divided concerning the suitability of ChatGPT and generative AI for operational integration and capacity to drive up profits via cost reduction, nearly 75% agree that this technology can be used for a range of core accounting functions, ushering in a profound shift in professional practice.
Tapping into new opportunities
At the heart of this unfolding debate is AI’s ideal role in our industry, as not all tax and accounting tasks are equally well-suited to automation. Moreover, early adopters – from ‘Big 4’ accounting firms to startups and SMEs – will simply not be able to unlock AI’s full operational potential without a clear strategic vision separating automation-friendly and human-dependent functions.
Over the past year, several fascinating studies have revealed AI’s relative strengths and weaknesses and should guide firms’ AI integration strategies. For example, BYU accounting professor David Wood conducted a research project testing ChatGPT’s performance in a range of tasks, notably finding that while the LLM proved capable in auditing, it fell roughly 30 percentage points short of his students in more complex, technical and maths-heavy problems, such as tax, financial and managerial assessment.
Similarly, Wharton School of Business professor Christian Terwiesch’s ChatGPT experiment discovered significant mathematical shortcomings while conversely establishing generative AI’s impressive competence in process analysis and operations management as well as “tedious” manual tasks including invoice processing, data entry and expense tabulation. Academia’s conclusions notably chime with the accounting industry’s anticipated AI-assisted transition from time-consuming, repetitive work to higher-value strategic and advisory work requiring uniquely-human communication, judgement and critical-thinking skills.
Crucially, AI-automated accounting is increasingly allowing professionals to generate and analyse vast swathes of data in real-time, informing faster, more accurate and more impactful decision-making and financial predictions needed to guide companies forward in turbulent times.
Looking at the horizon
Amid growing disaster narratives fuelled by each incredible advance of generative AI, the more tech-averse in our profession fearing impending AI-induced obsolescence should rest assured. The uncovered shortcomings of LLMs remind us that there will always be a need for human decision-making and complex accounting capabilities, as well as human oversight to correct AI’s occasional ‘hallucinatory’ inaccuracies and address any regulatory, security or ethics failures – from proprietary data leaks in the public cloud during LLM training to transparency and bias issues.
AI technologies should rather be seen as complementary tools, helping accounting firms to boost productivity and competitiveness while opening new possibilities to innovate, add value to client services and drive long-term business growth. While applying to the entire sector, AI’s potential is particularly crucial for small and medium practitioners (SMPs) with smaller budgets and less internal accounting capacity. Indeed, the accounting and finance industry faces a severe skills shortage for various reasons, with nearly three-quarters of European employers reporting recruitment issues in 2020 and nearly 20% claiming that this lack of talent was hindering key business objectives.
Across the pond, the US faces a similarly worrying shortage of trained accountants – particularly in audit work well-suited for AI automation – with roughly 75% of professionals at or near retirement age and firms unable to hire young people in sufficient numbers. Given the accounting industry’s historic reputation as a slow tech adopter and high-stress, long-hours working environment all of which are undermining its appeal to younger generations, firms should equally mobilise AI technologies to build an innovative image and enhance employee well-being through an efficiency-enabled workload reduction. In this space, AI software startup DataSnipper – which recently reached the unicorn milestone – is leading the way, offering the largest, mid-tier and smaller accounting businesses the financial capacity to retain current staff and automate a host of tedious tasks that fuel burnout and repel new recruits.
Increasingly displayed over the past year, AI’s vital role in the accounting industry is set to expand significantly as businesses’ growing interest in integrating this technology combines with the dizzying progress fuelled by OpenAI and its newest competitor on the scene, Mistral AI. With the right strategy and mindset, our profession and the wider finance industry will be able to attract the new talent and deliver the level of human-oriented service excellence needed, to stay competitive in the future.
