EY Tax Leaders Want AI Agents. Their Data Isn’t Ready.

EY’s 2026 survey of 1,600 finance and tax leaders found 77% expect AI agents to help, yet 71% remain early-stage and only 16% call their tax function highly effective with data.

EY survey finds tax teams keen on AI, but data lags
EY survey finds tax teams keen on AI, but data lagsAI-generated
By Elena Vasquez7 min read

Why it matters

  • EY’s 2026 survey covered 1,600 leaders across 30 jurisdictions and 22 industries.
  • 77% expect AI agents to benefit tax, up from 21% a year earlier, while 71% remain at nascent or exploratory maturity.
  • 64% cite insufficient AI-ready data as the biggest barrier; only 16% rate tax data effectiveness highly.
  • 83% value human-AI agent teams, 81% prioritize AI-related hiring, and 60% rank strategic thinking and problem-solving among the five critical future skills.
  • 79% are adapting to geopolitical developments, and 64% are not fully prepared for real-time tax demands.

By Elena Vasquez

Seventy-seven percent of tax and finance leaders expect AI agents to benefit the tax function, up from 21% a year earlier, yet 71% remain at a completely nascent or exploratory stage of maturity.

EY’s 2026 Tax and Finance Operations Survey, published on November 18, 2026, draws on responses from 1,600 Chief Financial Officers and senior finance and tax leaders across 30 jurisdictions and 22 industries.

The survey caught a stark split between expected returns and present capability. Interest in AI agents has risen sharply, according to EY, but most tax functions have barely moved beyond exploration. For tax departments, the stakes extend beyond software budgets, because EY also found pressure from geopolitical shifts and increasingly digital, immediate tax administration.

What is blocking AI in tax teams?

Data readiness emerged as the main obstacle in EY’s findings. The survey identified four related constraints:

  • 64% of tax and finance leaders identified insufficient AI-ready data as the biggest barrier to advancing AI initiatives.
  • Only 16% considered their tax function highly effective at accessing, organising, using and reusing data.
  • Among respondents whose function was not effective with data, 61% pointed to data silos and 52% cited poor data quality.
  • A further 43% said the inability to execute a sustainable data, AI and technology strategy was the biggest challenge to delivering their function’s purpose.

Together, these figures show where investment has not yet translated into operational capacity. EY says spending on AI has not resolved more basic operating problems inside finance and tax departments. Leaders must improve governance, data management and operating models while they adopt new tools. The appetite is clear. The supporting foundation is not.

Marna Ricker, Global Vice Chair-Tax at EY, said:

“Many organizations have already invested in AI. The challenge is transforming their data, governance and talent capabilities to ensure they get return on their AI investments.

As AI becomes embedded in business decision-making, ambition alone will not be enough.

Tax leaders are having to be deliberate in how their function creates value for the wider organization and have an opportunity to use trusted data, knowledge, AI, and human judgment to unlock capital.”

Ricker’s diagnosis matches the survey. Returns on AI spending depend on converting fragmented information into trusted data while strengthening governance and talent. Her answer links every part of that work instead of treating software deployment as a stand-alone fix.

For tax leaders, that creates a demanding sequence. They need usable data before they can deploy agents at scale. They need governance before those agents can support consequential decisions. And they need staff who can combine technical tools with human judgment. EY’s findings suggest that skipping any of those steps will leave ambition running ahead of execution.

What changes for tax teams and their workers?

EY’s survey treats AI adoption as workforce design, not merely a software purchase. Leaders expect people and agents to work together, while also changing what skills they recruit:

  • 83% said building a workforce in which humans work alongside AI agents is important.
  • 81% said hiring professionals with AI-related skills is a priority.
  • 60% placed strategic thinking and problem-solving among the five most critical skills for future tax professionals.

Respondents did not expect automation to remove the need for human involvement. The results instead place judgment and oversight at the center of increasingly automated processes. Human staff still matter, but EY’s survey indicates that their role will change alongside the systems they use.

Stuart Lang, Global Tax and Finance Operate Leader at EY, said:

“The future tax function won't separate compliance from insight.

Every filing and reconciliation generates trusted data and knowledge that, combined with AI and human judgment, can become a source of tax intelligence for the business.

Realizing that potential will require leaders to make deliberate choices about workforce and talent priorities, investing not only in technology but in people skills and culture.

Organisations that upskill their teams, build trusted data through strong governance and choose the right external providers will be better positioned to turn AI ambition into sustainable value.”

Lang’s model widens the role of tax beyond compliance. Under it, filings and reconciliations create reusable information for wider business decisions. That gives the 83% figure on human-and-agent teams a practical context. AI agents can handle parts of the process, but people still have to set priorities, exercise judgment and ensure the underlying information is trusted.

The staffing numbers support that view. EY found strong demand for AI-related hiring and a clear preference for combining technical skills with strategic thinking. The 60% finding on problem-solving also counters any assumption that efficiency gains will make experienced tax professionals unnecessary. Their work shifts toward directing systems, assessing outputs and connecting tax data to business needs.

Lang also placed external providers within that operating model. His recommendation reflects the survey’s finding that some companies already use outside support to manage demanding work and specialized requirements. The issue is not whether outside help exists. EY’s results show that companies need trusted data, suitable skills and careful provider choices before that help can deliver durable value.

What external pressures are tax teams facing?

Beyond AI adoption, EY’s survey captured broader business and policy pressures reshaping tax operations. The strongest finding was geopolitical: 79% of organisations said they were adapting their operations in response to geopolitical developments. That share indicates how much wider business and policy changes now shape tax strategy.

Another 64% said they were not fully prepared for real-time tax demands as tax administration becomes more digital and immediate. This readiness gap compounds the data problem. Teams already struggling with silos and poor-quality information must also prepare for faster, more digitally delivered tax processes.

Together, these pressures make the underlying operating model more important than another isolated AI pilot. Trusted data can support both routine filings and faster responses to external change. Governance determines whether that data remains usable. The survey portrays tax transformation as a business-wide discipline rather than a technology project confined to one team.

Some companies are also turning to external support. Among respondents already using co-sourcing, EY found:

  • 86% said co-sourcing gave them a greater ability to focus on higher-value activities.
  • 78% reported improved tax risk management.
  • 77% said it helped them navigate legislative and regulatory change more successfully.

These results give external providers a clear role in the market EY surveyed. Co-sourcing did not stand apart from AI readiness; it sat alongside the same demand for specialist capacity, trusted data and reliable execution. Lang’s reference to choosing the right external providers reflects that connection between technology, talent and outside support.

Can AI ambition become sustainable value?

EY’s central finding is a mismatch between what tax leaders expect from AI and what their functions can currently deliver. The debate has moved beyond whether agents will take on a larger role. The practical question now centers on whether data, governance and skills can support that role at scale.

That mismatch leaves several linked requirements for the next phase of tax AI:

  • Trusted data: 64% of leaders cite insufficient AI-ready data as their biggest barrier, while only 16% rate their tax function highly effective with data.
  • Governance and operating models: Ricker and Lang both connect returns with deliberate governance, data management and choices about how the function operates.
  • People and judgment: 83% value human-and-agent teams, 81% prioritize AI-related hiring, and 60% retain strategic thinking among the five most critical future skills.
  • External support: Among existing users of co-sourcing, 86% report more capacity for higher-value work, 78% report better tax risk management and 77% report greater success with legislative change.

For companies, EY’s findings point to sustained demand for AI agents, data capabilities and specialist support. The constraint lies in execution. Buyers may have moved faster than their operating foundations, but the survey provides no sign that appetite for adoption has faded.

For public authorities, the survey offers a private-sector readiness benchmark as tax administration becomes more digital and immediate. The 64% shortfall in real-time preparation matters because faster administration still depends on companies maintaining accurate, accessible and governed information.

Lang’s standard for sustainable value is concrete: upskill teams, build trusted data through strong governance and choose external providers carefully. Ricker’s standard is equally direct: ambition alone will not be enough. The survey suggests the next phase of tax AI will be decided by whether early-stage functions can turn trusted data, governance and human judgment into repeatable value, rather than simply adding more AI agents.

Source: itbrief.com.au

Share this article:

More from Elena Vasquez

Elena Vasquez

Show full bio

Market editor covering technology and product news in artificial intelligence and machine learning at AI In Context.

708 articles

Related articles

  1. Democrats See AI Crackdown as Path Back to Power in Congress
  2. Only 34% of Financial Services Content Is Connected, Iron Mountain Finds
  3. 51% of AI projects miss targets, yet 85% of firms raise budgets
  4. Productivity leads enterprise AI, but autonomy remains scarce
  5. Meta Saved $3.9 Billion in 2025 by Calling AI Data Centers Experiments

« Previous articleNext article »