AI adoption rises, but 93% of finance professionals question AI-generated insights: Report
New Delhi, July 16
Even as finance teams increasingly adopt artificial intelligence to improve business decision-making, 93 per cent of finance professionals remain concerned about the integrity and verifiability of AI-generated insights, according to a joint report by ACCA and Chartered Accountants Australia and New Zealand.
The report titled - "Enabling Finance Insight: Bridging Skills and Data Gaps for AI-enabled Finance", based on a global survey of 1,600 finance professionals, said concerns stem from issues such as AI hallucinations, inaccuracies, incomplete datasets, lack of transparency and bias, highlighting the need for stronger governance and upskilling as AI adoption gathers pace.
The report said the finance function is moving beyond its traditional role of historical reporting, with organisations increasingly expecting finance teams to provide forward-looking business insights. It noted that "the finance function stands at an unmissable opportunity," as stakeholders demand "proactive leadership - requiring finance to evolve from a retrospective reporting engine into a strategic enabler of enterprise-wide insight."
According to the report, while AI is becoming a core part of finance's analytical toolkit, organisations must deploy the technology strategically to create business value rather than simply automate existing processes. "AI is becoming a fundamental component of finance's analytical toolkit - finance leaders must strategically deploy these technologies to generate value, not merely automate existing inefficiencies," it said.
The study also found that finance teams are increasingly relying on real-time operational data and AI-powered analysis. More than 60 per cent of finance teams have increased their use of real-time operational data over the past two years, while the use of internal text data such as meeting transcripts, contracts and documents is also growing as generative AI tools become embedded in day-to-day work.
However, the report said poor data quality, skills shortages and difficulties in integrating multiple data sources remain the biggest barriers to using AI effectively. Data quality issues and lack of appropriate skills were each cited by 42 per cent of respondents, while 40 per cent pointed to the challenge of integrating data from multiple sources.
It also highlighted a widening skills gap as AI adoption accelerates. According to the survey, 72 per cent of respondents reported having only basic or no generative AI skills, although 41 per cent said they are pursuing AI training and upskilling on their own.
ACCA Chief Executive Helen Brand said finance leaders must focus on governance alongside technology adoption. "CFOs and finance teams need to lead in the responsible adoption of AI across organisations, ensuring robust training and governance is in place. Critical thinking, sceptical validation and an ethical approach is vital," she said.
The report concluded that finance teams are uniquely positioned to lead responsible AI adoption because of their role in governance, data stewardship and performance measurement. It recommended greater investment in structured learning, stronger collaboration with IT and data teams, and improved data governance to ensure AI delivers trusted insights and measurable business value.
— ANI
Reader Comments
As someone working in a Big 4 in Gurgaon, I can totally relate. We get constant pressure to 'use AI' but half the time the models hallucinate numbers. Last month, an AI tool invented an entire revenue projection based on outdated tax laws. Thank God for professional scepticism—my two days of manual verification saved us from presenting nonsense to a client. The 93% statistic doesn't surprise me one bit. We need human oversight, not blind automation.
I'm a CFO in a mid-sized American firm, and we had the exact same debate last quarter. The report is spot-on: AI is useless if your data is messy. We spent $200k on an AI platform only to realize our sales data had duplicate entries and missing fields. The real transformation is in data hygiene and training people to ask the right questions, not in buying more software. Glad to see ACCA flagging this globally.
Yaar, but we can't ignore the upside. My team uses AI for basic expense classification and vendor reconciliation, and it saves us 10 hours a week. The problem is when management expects AI to replace critical thinking. Finance is about judgment, not just number crunching. If we upskill properly—like learning how to validate AI outputs—we could actually be more strategic. The 41% who are self-learning? That's the right spirit! 🇮🇳
CA from New Zealand here—this report is exactly what we needed. The 'move fast and break things' mindset doesn't work in finance where one hallucinated number can cost a company millions. I'm impressed that Indian firms are adopting real-time data (60% figure is higher than here), but the skills gap is alarming. I spent my own money on a Python for Finance course. If we don't invest in people, we'll end up with broken AI systems and angry auditors.
We welcome thoughtful discussions from our readers. Please keep comments respectful and on-topic.