The rise of AI in professional services is pushing law firms, consulting partners, and accounting groups to redesign client work around trusted business outcomes rather than billable time [1]. Over the past year, GenAI tool adoption doubled in the advisory sector, yet firm-level revenue returns are still uneven across practice groups [2]. While knowledge workers adopt AI assistants on their own, company-wide adoption faces hurdles in billing models, partner governance, and client privacy rules [3]. The Thomson Reuters Institute surveyed more than 1,500 experts in its 2026 report, finding that 40% of firms now use GenAI tools in daily work, up from 22% last year [2].
AI in Professional Services Redefines Daily Workflows
Adoption in advisory firms has grown fast over the past two years, altering how teams handle routine inquiries, document reviews, and background client research [3]. A 2025 study from Accenture found that 78% of advisory staff used GenAI tools within six months of public release, while the 2026 Thomson Reuters Institute report showed that more than 80% of regular users engage with smart systems every week, with 66% expressing optimism about their daily workplace role [2]. Yet this shift reveals a split between personal tool use and firm-wide payoff. Only 19% of surveyed firms report having no plans for adoption, while 56% of chief leaders in PwC’s 29th Global CEO Survey reported no significant cash benefits from their current AI rollouts [1].
Speaking on SiliconANGLE’s livestreaming show theCUBE, software sector advisor Matt Cook from PwC U.K. joined Certinia leader Prasad Narasimhan Sulur to discuss why adding more tools doesn’t bring higher revenue. Cook said that winning firms redesign value creation itself, rather than simply speeding up older manual tasks. “Where we’re seeing AI leaders win is they’re not simply deploying better tools. They’re actually redesigning how that value is created,” Cook said. Sulur pointed to software coding, explaining that faster coding shortens writing cycles but doesn’t eliminate delays in testing, system checks, or final product release [1].
Lasting progress depends on linking top direction with team capability, because top-down mandates fail when employees lack the skills to apply new tools to daily client problems, while combining executive leadership with bottom-up proficiency moves firms forward. Sulur noted that both sides must align [1]. In an academic review conducted at the University of Wollongong, lead researcher Carl Bezuidenhout synthesised findings from 75 peer-reviewed studies in the Journal of Information & Knowledge Management [4]. The review concluded that smart systems are shifting professional services firms (PSFs) toward data-intensive teamwork [5].
The Economic Paradox of the Billable Hour
Efficiency gains create a revenue tension that other sectors don’t share, since speed directly undermines total fees whenever pricing depends on hours billed rather than delivered value. For decades, law firms and advisors billed clients by calculating hours spent on tasks. When an automated tool finishes a 16-hour research review in four hours, an hourly model cuts firm revenue by 75%, penalizing the firm for its own speed. Deloitte’s 2025 benchmark showed that firms keeping time-based billing grew revenue by 2.1% annually, while firms using value-based pricing grew by 8.7%. The margin difference reached 6.6 percentage points [3].
To preserve margins, partners are decoupling client fees from hours worked by packaging advice into fixed-fee retainers, outcome milestones, and subscription advisory products that reward rapid delivery. In legal work, Allen & Overy paired its rollout of Harvey with fixed-fee deals, letting the firm turn faster execution into higher profit. Digital shifts in other advisory markets show a comparable dynamic, including research examining dehumanizing rental platform algorithms where software speed eroded client trust when automated checks replaced human consideration [3].
Rethinking fees also reshapes project staffing, altering the legacy leverage model where junior staff handle document checks and legal research that account for up to 70% of total billed hours. AI review can cut paralegal task volume by 45%, based on Altman Weil’s 2025 law firm survey. Firms that fail to reshape junior roles face 28% higher staff attrition, while firms training staff to supervise AI systems build stronger practices that retain top young talent [3].
How Professional Services AI Demands Human Judgment
As models take over repetitive analysis, the value of seasoned human judgment grows because algorithms cannot accept legal or cash liability when advice proves wrong. Cook observed that AI software accelerates evidence collection and scenario testing, but it can’t assume professional responsibility. “AI can accelerate evidence gathering and scenario development, but it doesn’t carry professional accountability,” he said [1]. In advisory matters, clients pay for accountability under uncertainty rather than raw computational speed [7].
Treehouse Innovation identified three core domains where human expertise must remain the main product: judgment under uncertainty, ethical choices, and client trust. When novel disputes arise or ethical boundaries blur, algorithmic models can’t weigh human consequences. AI tools excel at processing files, finding patterns, and producing initial drafts for review. That division leaves experts free to focus on strategic guidance and negotiation [7].
Changing workflow models also changes hiring expectations, as entry-level staff take on duties that once required years of senior apprenticeship. PwC research shows that junior roles exposed to automation increasingly require skills once expected only of senior leaders, including decision-making and client communication. Cook noted that when machines generate more drafts than people can read, human judgment becomes the true differentiator. “When AI can produce more analysis than a human can read, the differentiator’s not more content, it’s better judgment,” Cook said [1].
Overcoming Governance Friction and Knowledge Silos
Rolling out shared tools meets steep friction in partner-led firms, where individual partners hold near-total autonomy over client accounts, staffing, and software choices across their own distinct practice groups. McKinsey’s 2025 survey of 200 partnerships found that 63% lacked a written firm-wide AI strategy, leaving practice groups to run separate experiments that duplicated software costs and fragmented data. ALM Intelligence quantified this overhead: the average law partnership needed 4.7 committee meetings to approve an enterprise tool, compared to 1.2 approvals in in-house legal teams [3].
Unorganized files constitute a firm’s greatest asset, yet much of that value sits locked in email archives, personal inboxes, and local storage drives. An IDC study found that typical advisory firms can structurally access only 15% of their expert assets through existing systems. Structuring past project documents through retrieval-augmented generation (RAG) unlocks institutional memory by letting search tools extract verified precedents from previous client engagements. Between 2023 and 2025, PwC invested USD 1 billion in knowledge systems to make historical casework accessible to every team, helping consultants build proposals 40% faster while lifting win rates by 15% [3].
Data protection also intersects with ethical obligations around bias and client privacy, since attorney-client privilege bars firms from feeding sensitive records into public systems [3]. In addition, skewed datasets can taint audit findings, mirroring challenges described in research on healthcare AI bias in training data where unbalanced records warped diagnostic models [4]. Safe deployment requires private environments, tiered permissions, and firm agreements prohibiting vendor training on submitted text [3].
Industry Tools Win for AI in Professional Services
Broad consumer chatbots are yielding to domain-specific platforms designed for regulated environments, where general-purpose tools lack required compliance standards and fail to guarantee client data safety. Usage of specialized industry tools rose 14 percentage points in 2026 as firms prioritized reliability. In personal injury litigation, Thomson Reuters integrated CoCounsel Legal with Supio to manage case files. Supio organizes complex medical records, hospital bills, and damages narratives into structured ledgers, while CoCounsel tests those claims against authoritative legal research and precedent to build defensible courtroom arguments [6].
Despite broad use, tracking financial returns is still an industry blind spot because most leadership teams lack clear metrics for technological performance. Only 18% of firms measure return on investment from artificial intelligence, and 40% of experts don’t know if impact is tracked at all. Measuring firms usually focus on internal hours saved rather than revenue gains. Mike Abbott, head of the Thomson Reuters Institute, said that 2026 represents a strategic phase where firms must connect tools directly to business planning [2].
The coming decade centers on agentic systems that run complex research tasks on their own, fundamentally changing project execution timelines across the entire advisory and legal landscape. Thomson Reuters found that 77% of professionals expect agentic AI to be central to daily workflows by 2030, though only 15% use it in production today, leaving massive room for expansion [2]. While 48% of in-house legal teams support agentic automation, only 17% feel comfortable letting systems provide legal advice [6]. Winning firms will pair computational power with human oversight, keeping responsibility in human hands [1].
- ONLINE NEWS Knight, C. (2026). Human judgment reshapes AI services. SiliconANGLE. [Article Link]
- REPORT Thomson Reuters Institute. (2026). 2026 AI in Professional Services Report. Thomson Reuters. [Article Link]
- WEBSITE Pucek, B. (2026). AI in Professional Services: Complete 2026 Guide. The Thinking Company. [Article Link]
- ACADEMIC JOURNAL Bezuidenhout, C., Abbas, R., Mehmet, M., & Heffernan, T. (2025). Artificial Intelligence in Professional Services: A Systematic Review and Foundational Baseline for Future Research. Journal of Information & Knowledge Management, 24(02). [Article Link]
- ONLINE NEWS Science Matters. (2026). AI is Reshaping Professional Expertise and Business Models in Law, Healthcare and Finance. The Science Matters. [Article Link]
- WEBSITE Thomson Reuters. (2026). Highlights from the 2026 AI in Professional Services Report and What It Means for Legal Teams. Thomson Reuters Law Blog. [Article Link]
- WEBSITE Treehouse Innovation. (2026). AI Opportunities for Professional Services. Treehouse Innovation. [Article Link]
