AI is no longer the question in professional services, at least at the individual level. A new Thomson Reuters report argues the real problem is operational, where organizations struggle to turn widespread AI usage into dependable, governed outputs that clients can trust.
In its Future of Professionals 2026 report, Thomson Reuters surveyed 1,816 professionals across law, tax, audit, accounting, compliance, risk, and global trade. The results point to an “execution gap” between AI ambition and day-to-day delivery, and warn that the downstream effects are becoming measurable in both talent and client relationships.
AI usage is rising, but value delivery is lagging
According to the report, 74% of professionals are using AI tools at least weekly. However, that familiarity is not translating into business outcomes. Thomson Reuters says 91% of respondents believe their organizations are falling short of what AI can deliver.
The mismatch appears in several ways. About 35% say AI ambitions do not show up in their day-to-day work, while nearly one in five report that their organization still lacks a clear AI strategy. Together, these findings suggest many firms are treating AI as an adoption exercise rather than a transformation one, where processes, controls, and delivery standards are updated to match the technology’s capabilities.
Revenue at risk as clients reassess providers
The report frames the economic cost of that gap as a client issue. It says 78% of corporate clients consider AI-enabled quality improvements essential or very important. Yet only 6% of professionals believe most providers are delivering those improvements.
That disconnect is leading to customer scrutiny. Thomson Reuters reports that within 12 months, 32% of corporate clients will reconsider provider relationships, and that a third of those clients say the reassessment puts more than $1 million of annual work at risk. Thomson Reuters estimates that, when applied to the U.S. legal and CPA markets, this equates to roughly $143 billion in revenue under active reconsideration.
While the figure is an estimate derived from survey responses, the underlying implication for firms is clear: clients are increasingly benchmarking service quality against AI-enabled outcomes, not just cost or speed.
Shadow AI creates governance and confidentiality concerns
One of the report’s most operationally urgent findings is that unsanctioned tool usage is becoming common. Thomson Reuters says about one-third of lawyers, accountants, and compliance professionals use AI tools that their organization has not approved. The rate rises to 41% among respondents who say their organization is moving too slowly on AI.
This matters because the work in these industries is typically tied to liability, confidentiality, regulatory obligations, and evidentiary standards. The report indicates most professionals see governance as non-negotiable. It reports that 96% say AI must safeguard confidential data, 94% require outputs supported by authoritative and verified content, and 90% want outputs they can explain and defend.
However, access does not always match these requirements. The report states that 41% of respondents lack access to professional-grade AI tools that meet the stated confidentiality and quality expectations. The practical risk is that employees may turn to general-purpose AI when sanctioned options are unavailable or fail to fit established workflows, creating “invisible” risk that organizations cannot easily monitor.
Talent retention pressure emerges when AI delivery disappoints
Professional services firms rely on expertise-intensive workforces, so technology gaps can become talent risks quickly. Thomson Reuters reports that one in four professionals experiencing a gap between what AI can do and what their organization delivers are considering leaving within two years. It also notes that 13% would consider leaving within 12 months.
At the leadership level, the report highlights a perception gap. While 24% of professionals foresee possible departures, almost half of senior leaders believe meaningful talent pressure is at least three years away. That timing difference can be costly if firms do not adjust training, delivery expectations, and technology access before the labor market responds.
The report also suggests hiring and role acceptance are tied to AI enablement. It says 62% of professionals would consider access to professional-grade AI as a factor in accepting a new role. For those already using such tools, nearly one in three would turn down a role without them.
What “accountability” means for AI in regulated work
The survey positions the issue as more than tool adoption. It argues that in areas where outputs influence legal judgments, regulatory filings, or client advice, the standard for AI performance must be higher than “helpful” or “good enough.”
Thomson Reuters links this to a concept it calls “Fiduciary Grade AI,” which it says is designed to be verifiable and supportable, built on authoritative domain content, and paired with privacy and security controls and human support. The report’s role, in this framing, is to emphasize that AI implementation in liability-heavy professions requires accountability mechanisms, not just model access.
Even without accepting the company’s specific branding, the report’s broader message aligns with what many regulated industries are already learning: governance, data handling, output traceability, and workflow integration are central to adoption.
Implications for law, tax, audit, and accounting leaders
For firms planning AI roadmaps, the report points to several priorities implied by the results:
- Operationalize AI, not just pilot it. Respondents report ambitions are not reaching daily workflows, which indicates delivery gaps in process design and change management.
- Close the gap between requirements and tool access. With a large share lacking professional-grade options, organizations may inadvertently push employees toward unsanctioned tools.
- Measure AI-enabled quality improvements. Client expectations are rising, and perceived delivery shortfalls are leading to provider reassessments.
- Treat talent risk as a near-term metric. Professionals contemplating departure suggest AI enablement is becoming part of the employment value proposition.
As AI becomes a baseline expectation across professional workflows, firms will likely be judged less on whether employees use AI, and more on whether organizations can govern it, prove its value, and support it consistently at scale.
