Thomson Reuters report says AI execution gap challenges KPMG and peers

AI is no longer the problem for professional services firms. Execution is. Thomson Reuters said 74% of the more than 1,800 professionals it surveyed across law, tax, audit, accounting, compliance, risk and global trade use AI tools several times a week, while 44% rely on them multiple times a day. Yet 91% said their organizations are falling short of what AI can deliver, a gap that lands squarely on firms like KPMG, where speed alone is not enough unless the output is reviewable, defensible and client-ready.

The stakes are sharpest in audit and tax, where about one third of lawyers, accountants and compliance professionals said they are using unsanctioned AI. That creates obvious confidentiality and model-risk issues, but it also points to a management problem: if people are already reaching for their own tools, leadership has not made sanctioned options fast or useful enough. Thomson Reuters said as much as $143 billion in U.S. client revenue could be at risk if firms fail to implement AI effectively, and that 24% of talent could leave within two years if they do not see AI value in their work. For KPMG staff, that turns governance, training and use-case ownership into career skills, not side topics.

The client side is just as unforgiving. Thomson Reuters said 78% of clients want AI-enabled quality improvements, but only 6% believe most providers are delivering them. That helps explain why the profession is moving past pilots and into operating discipline, where partner buy-in, workflow redesign and quality control decide whether AI changes billable work or just adds another layer of experimentation. Firms that can show how AI improves delivery without weakening assurance are the ones most likely to protect margins, keep people engaged and hold onto client trust.

KPMG has been making the same shift in its own messaging. In April 2025, it said it was accelerating AI integration into KPMG Clara, its global smart audit platform, with AI agents meant to support more than 95,000 auditors globally while keeping a human-in-the-loop model. In May 2025, it launched AI Trust services built around KPMG’s Trusted AI framework and ServiceNow AI Control Tower, with a focus on reliability, accountability, transparency, readiness, compliance, security and risk transformation. KPMG’s 2025 global trust-in-AI study, which surveyed more than 48,000 people in 47 countries with the University of Melbourne, pointed in the same direction: the firms that win this race will be the ones that can operationalize trust, not just talk about adoption.