AI Adoption Is High. AI Impact Is Low.
The 2026 Intuit Accountant Technology Survey of 725 professionals shows strong adoption numbers: 86% use AI for at least one firm operation. The top uses are invoicing and payments (53%), client communication (50%), and managing client portfolios. Firms are clearly investing.
But the same survey found that 90% report burnout as a significant issue, with 25% pointing to fragmented data as a key driver. If AI is this widely adopted and burnout hasn't dropped, something structural is wrong. The tools are faster. The underlying workflow hasn't changed.
The Audit Evidence Problem
Audit teams feel this most acutely. Every engagement generates hundreds of evidence requests. Bank confirmations, signed representations, board minutes, trial balances, lease schedules, HR records. Each item involves a back-and-forth between the audit team and the client's finance function.
Wolters Kluwer's 2026 survey of 4,214 internal audit professionals found that AI adoption in internal audit will double to 80% by 2026. But 46% said AI adoption in their audit function was lower than in other areas of the business. Audit is behind, not because the technology doesn't exist, but because the evidence collection process isn't structured enough for AI to work with.
AI Needs Structure to Deliver Value
When audit evidence lives in email threads, shared drives, and spreadsheets with no standardized format, AI can't meaningfully process it. An AI tool can draft a follow-up email faster. It can't tell you which of your 200 evidence requests are complete, which are partially complete, and which haven't been touched, if that information is scattered across three inboxes and a shared folder.
Wolters Kluwer's analysis is direct: technology adoption isn't just about buying tools. It's about integrating them. 74% of firms anticipate seeing impacts from the effort to keep up with tech, even as cloud and AI usage accelerate. The problem is integration, not adoption.
Structure First, Intelligence Second
The audit firms seeing real efficiency gains from AI share a pattern: they structured their evidence collection process before they added intelligence to it. When every document request has an owner, a deadline, and visible completion status in a shared environment, the AI has clean data to work with. Auto-classification works. Follow-up prioritization works. Status dashboards reflect reality.
The firms that added AI on top of email-based evidence collection just got faster at the wrong things.
See how audit firms are building structured evidence collection that makes AI actually useful.
