A Thomson Reuters survey of 1,816 professionals across legal, tax, audit, accounting, compliance, risk, and global trade indicates that AI adoption in professional sectors remains slow and chaotic. The 2026 report, conducted in March – April 2026 across 62 countries, reveals that one-third of surveyed professionals are using unapproved 'shadow' AI, defined as tools not officially sanctioned by an organization's tech stack and lacking appropriate security or data barriers. This figure climbs to 41% among professionals who perceive their organizations as adopting AI too slowly.

Furthermore, 25% of respondents are considering leaving their jobs within the next two years due to their firm’s failure to provide adequate AI tools. Despite these figures, nearly 50% of senior leadership believe that significant talent pressure related to AI adoption is still at least three years away. Steve Hasker, President and CEO of Thomson Reuters, noted that the standard for AI in high-liability professions must be significantly higher, advocating for 'Fiduciary-Grade AI' that can be verified and trusted in sensitive contexts such as legal judgments or regulatory filings.

The survey highlights a widening gap between employee expectations and corporate responsiveness. While many workers are eager to leverage AI for routine tasks, organizations that lag in providing secure enterprise-grade tools risk data security threats and reduced staff morale. The findings suggest that while entities like Big Law firms and Fortune 500 companies are often more proactive, the broader professional economy struggles to keep pace, creating hazardous work behaviors. The report concludes that AI integration is deepening monthly, making it imperative for firms to establish formal training and economic frameworks to support the use of secure, approved technological solutions.

A Thomson Reuters survey of 1,816 professionals across legal, tax, audit, accounting, compliance, risk, and global trade indicates that AI adoption in professional sectors remains slow and chaotic. The 2026 report, conducted in March – April 2026 across 62 countries, reveals that one-third of surveyed professionals are using unapproved 'shadow' AI, defined as tools not officially sanctioned by an organization's tech stack and lacking appropriate security or data barriers. This figure climbs to 41% among professionals who perceive their organizations as adopting AI too slowly.

Furthermore, 25% of respondents are considering leaving their jobs within the next two years due to their firm’s failure to provide adequate AI tools. Despite these figures, nearly 50% of senior leadership believe that significant talent pressure related to AI adoption is still at least three years away. Steve Hasker, President and CEO of Thomson Reuters, noted that the standard for AI in high-liability professions must be significantly higher, advocating for 'Fiduciary-Grade AI' that can be verified and trusted in sensitive contexts such as legal judgments or regulatory filings.

The survey highlights a widening gap between employee expectations and corporate responsiveness. While many workers are eager to leverage AI for routine tasks, organizations that lag in providing secure enterprise-grade tools risk data security threats and reduced staff morale. The findings suggest that while entities like Big Law firms and Fortune 500 companies are often more proactive, the broader professional economy struggles to keep pace, creating hazardous work behaviors. The report concludes that AI integration is deepening monthly, making it imperative for firms to establish formal training and economic frameworks to support the use of secure, approved technological solutions.