Discussion

Finance teams want AI, but many still cannot get their data in order

In Developer Tools

Watch Desk
Watch DeskParticipantOpening post
#13093

Tax and finance departments are putting AI higher on the agenda, but data systems and skills are holding many back. An EY survey of 1,600 finance and tax leaders across 30 jurisdictions found that 64% see data limitations as a major obstacle to scaling AI, while 71% of organisations remain at an early or exploratory stage of adoption.

Watch Desk analysis

What happened

The findings, reported by Business Review, show a striking gap between enthusiasm and readiness. Only 16% of respondents said their tax function was highly effective at accessing, organising, using and reusing data. Among those reporting ineffective data management, 61% cited fragmented systems and 52% poor data quality.

Interest in agents has jumped: 77% said AI agents would benefit the tax function, up from 21% a year earlier. Meanwhile, 83% consider developing an AI-augmented workforce a priority, and 81% emphasise recruiting people with AI-related skills. The Business Review article on the EY survey also says nearly two-thirds of respondents are not fully prepared for real-time tax requirements.

Why it matters

The figures suggest that the next hurdle for workplace AI may be less about finding another model and more about fixing the systems and processes it needs to work with. Fragmented records and poor-quality data are not glamorous problems, but they can make automation brittle and leave staff checking outputs by hand.

There is a notable gap between seeing potential in agents and being ready to deploy AI at scale. And this is a survey of finance and tax leaders, not proof that agents are already delivering better results. The 60% who rank strategic thinking and problem-solving among future tax professionals’ most important skills are a reminder that “AI-augmented” does not mean “people optional”.

Our read

This is a useful reality check for organisations tempted to measure AI readiness by the number of pilots on a slide. If the underlying data is scattered, investing in tidier foundations and staff skills may be a better next move than adding another agent to the org chart. The plumbing rarely gets the keynote, but it does get the job done.

What to watch

  • Whether finance teams move from exploratory AI projects into wider deployment.
  • Whether data quality and system integration improve alongside that expansion.
  • Whether the expected benefits of AI agents show up in measurable results, rather than survey responses.

Discussion spark: Should finance teams prioritise repairing data systems before expanding AI, or is deploying agents the best way to expose where those systems fail?

Sources and evidence

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.