AI was supposed to make accountants more strategic. Instead, it’s creating a skills crisis that most firms haven’t noticed yet.
Here’s what’s happening: 80% of accounting firms are seeing surging demand for advisory services – financial planning, business strategy, the high-value work everyone wants to do.
At the same time, AI is automating the compliance work that used to teach accountants how to actually deliver that advisory work. The technical immersion that built expertise – spotting anomalies in a client’s R&D spend, flagging misaligned payroll patterns – isn’t happening as routinely as it did in the past.
Experienced accountants will be fine. They’ve done enough manual compliance work to retain that intuition. But the next generation? They’re learning to be advisors without ever building the compliance foundations that make good advisory possible.
And with the global accounting advisory market projected to grow from $101.62 billion in 2024 to $165.15 billion by 2034, firms can’t afford to get this wrong.
Why automation creates shallow expertise
The problem isn’t automation itself. It’s what automation removes from the learning process.
When an accountant manually prepares a corporation tax return – spending 30 hours deep in the numbers – they’re not just processing data, they’re developing pattern recognition. They’re building an intuitive understanding of how businesses work. They’re learning to spot the narrative behind the numbers.
95% of accountants say technology has reduced time spent on compliance tasks. That’s the win everyone celebrates. But what nobody’s measuring is that those same accountants no longer need to understand client details the way they once did. They can’t go as deep on certain fields because the software does it for them.
Technical depth and advisory capability are different skills, but one depends on the other. Firms are discovering they can’t simply automate compliance and expect advisory capabilities to emerge.
Strategic thinking, consultative approaches, business empathy – these require deliberate development. And most firms haven’t built the training infrastructure to replace what automation took away.
The divide that’s already forming
This skills gap is splitting the profession into distinct tiers.
At the top are firms that saw this coming. They’re embedding AI directly into compliance workflows – not as bolt-on tools, but as native features that automate reconciliations, flag issues and surface insights automatically.
More importantly, they’re using that efficiency to democratise advisory work. Software that scaffolds client conversations is letting junior staff access knowledge that used to require years of experience. These firms are building advisors faster than ever before.
In the middle are firms still trying to figure out how to use AI effectively. They’re investing, but often in the wrong places – treating ChatGPT as a research assistant rather than embedding automation where compliance actually happens.
And at the bottom are firms that haven’t started. Often led by partners nearing retirement, these practices face the perfect storm: new hires aren’t developing deep expertise and experienced candidates are increasingly hard to find.
Why chatbots aren’t the answer
But even in those more ‘progressive’ firms, here’s where they’re at risk of getting their AI strategy wrong.
Public LLMs like ChatGPT are impressive. They summarise research brilliantly. They explain concepts clearly. They answer questions conversationally. But they can’t do the complex calculations and data-secure quantitative analysis that accounting demands.
More importantly, they don’t solve the core problem – building advisory strength on compliance strength.
Firms can only free up capacity for strategic client work if they dramatically reduce time-to-compliance. That doesn’t happen with research assistants. It happens when automation is embedded directly where the work occurs – in the bookkeeping software, the compliance stack, the production workflows.
The firms seeing genuine ROI aren’t prompting chatbots. They’re adopting AI-native features that actively automate processes within their existing systems. When compliance is automated at the source, insights flow naturally into advisory conversations.
Accountants get richer talking points and better context without changing how they work.
What separates winners from laggards
The firms that will dominate are the ones connecting three things.
First, they’ve made a decisive choice about direction. Some firms are leaning heavily into advisory, using compliance as the foundation. Others are positioning themselves as compliance specialists who offer advisory when clients need it. Both are valid paths, but they require completely different investments in training, hiring and technology. The firms struggling most are trying to split their focus evenly without committing resources to either direction.
Second, they recognise that AI adoption is fundamentally a workforce transformation, not a technology implementation. Software alone doesn’t create results. Training programmes, governance frameworks and cultural change management do.
Third, they’re intentional about developing advisory capabilities. That might mean pairing junior staff with senior advisors, building structured training programmes or deploying software that surfaces client insights for conversations.
Whatever the approach, they’re not leaving skill development to chance.
The opportunity disguised as a crisis
The next 18 months will separate the accounting firms that understand this moment from those still treating AI as an efficiency play.
Winners will build what we might call ‘AI-native advisory practices’ – firms where compliance automation and skills development are designed together, not bolted on separately. They’ll stop asking ‘how do we use AI?’ and start asking ‘how do we build accountants who can deliver world-class advisory when compliance is more automated?’
That’s a fundamentally different question. It requires different technology choices (workflow-embedded over chatbot experiments), different training investments (advisory skills over technical depth) and different business models (value-based over time-based).
The expertise crisis is already here. But for firms willing to rebuild how they develop talent, it’s also the biggest opportunity accounting has seen in decades.
