AI's Double Disruption: Professional Services Under Pressure

Generative AI is rapidly becoming part of day-to-day work in consulting, legal, accounting and advisory firms. One recent survey found 66% of practitioners support using generative AI in daily work and feel optimistic about its impact ([1]). In the legal sector alone, 79% of professionals now report using AI tools – a fourfold increase from just 19% in 2023 ([2]). Tools like AI contract review can cut document analysis time by up to 85% while achieving around 95% accuracy, compared to roughly 80% accuracy with manual review ([3]). These efficiency gains are freeing up senior experts from routine tasks and speeding up project delivery across the board.

Firms are investing heavily to embed AI throughout their organizations. Consulting giants have rolled out internal AI assistants to support their armies of consultants. McKinsey, for example, deployed its "Lilli" generative chatbot to more than 7,000 consultants, and by late 2025 about 72% of McKinsey’s 45,000 employees were actively using it ([4]). The Big Four accounting firms have similarly gone all-in: EY has 150 AI "agents" assisting 80,000 of its tax professionals ([5]), and KPMG has committed $2 billion over five years to develop AI systems in pursuit of an estimated $12 billion in new AI-driven revenue ([6]). These are not mere experiments – the Big Four have already integrated AI across their core service lines in production at scale ([7]).

This push is giving rise to new roles and training programs focused on AI. In a first for Big Law, Pillsbury announced the hire of a Chief Artificial Intelligence Officer to lead its firm-wide AI strategy ([8]). The veteran legal tech expert will leverage AI tools to streamline workflows, improve analysis, accelerate knowledge access, and support high-quality legal work while ensuring that legal judgment, strategic thinking, and client counseling remain firmly in the hands of experienced lawyers ([9]). It’s an explicit acknowledgment that AI can handle the heavy lifting, but expert human oversight is still critical. At another major firm, Haynes Boone, working with generative AI is now considered a core lawyering competency ([10]). The firm has rolled out a continually evolving mix of on-demand videos and hands-on workshops (with support from legal tech trainer Hotshot) to ensure every attorney is comfortable using AI in practice.

Still, rapid adoption has not been without missteps. In their eagerness to tout AI prowess, some firms have stumbled on quality control. In one notable incident, an EY report was found rife with fake citations and 'AI slop' – apparently the product of overzealous use of generative tools without proper oversight ([11]). The embarrassment highlighted the need for robust review processes and human judgment even as firms embrace automation. It’s a cautionary tale: AI can boost efficiency, but unchecked it can also produce errors that damage credibility.

Just as professional firms are using AI to reinvent their work, their clients are doing the same – potentially reducing their reliance on outside experts. In corporate law, for instance, an ACC/Everlaw study found generative AI adoption by legal departments more than doubled from 23% to 52% within a year ([1]) – and nearly two-thirds of in-house counsel (64%) now expect to rely less on outside law firms as a result ([2]). This rapid self-sufficiency is already shifting the balance of power and spend in legal services, pressuring traditional firms to prove their value beyond routine document review and research.

Similar patterns are emerging across other industries. Many tech-savvy companies have started building internal AI capabilities rather than paying multimillion-dollar consulting fees ([3]). The logic is hard to ignore – as one industry commentator put it, 'Why hire a big firm’s junior team to churn through data if an off-the-shelf AI and a couple of your own analysts can do it faster and cheaper?' ([4]). This “do-it-yourself” AI approach is nascent but growing, as organizations realize they can keep institutional knowledge in-house and use affordable AI tools to handle tasks that used to require outside contractors.

Meanwhile, the major AI technology companies themselves are moving into the advisory arena. Both OpenAI and Anthropic have launched new business units to provide AI consulting services directly to enterprises ([5]). These moves are backed by billions in capital and aim to capture a slice of the $1 trillion global consulting and tech services market. A senior Big Four insider remarked that such ventures are 'definitely a direct play against the traditional consulting firms' ([6]) – essentially, the AI labs are now competing with the consultants.

The tech entrants are bringing formidable capabilities that encroach on work once done exclusively by human advisors. OpenAI’s new Deployment Company, for example, is embedding 'co-pilot' AI agents that can draft strategy documents, analyze financial statements, and even conduct audit-style checks autonomously ([7]). OpenAI jump-started this offering by acquiring a 150-person AI consultancy in the UK ([8]). Clients have noticed the potential – and many prefer it. One consulting source noted that customers are 'demanding more work is done by AI bots' for lower-level tasks and that 'they don’t want a traditional MBA grad coming in with a slide deck and telling you how to cut costs' anymore ([9]). In turn, even the largest incumbent firms are, as this source put it, having to 'completely rewrite how they operate' ([10]). Between client in-housing and direct competition from AI providers, the message to legacy professional service firms is clear: adapt quickly or risk irrelevance.

Perhaps the most profound changes are unfolding in how professional services are priced – and in what "expertise" really means in an AI-enabled world. Many industry experts predict the traditional leveraged pyramid structure of firms (with a broad base of junior billers supporting a few partners at the top) will give way to a leaner model. AI-first firms are already moving toward a slimmed-down 'obelisk' organization with far fewer junior staff, heavy use of technology, and new fee models that move away from billing by the hour ([1]). Clients are certainly receptive: a recent benchmark found 67% of small businesses now prefer fixed or outcome-based pricing for initial AI projects instead of open-ended hourly rates ([2]). As AI makes routine analysis faster and cheaper, buyers increasingly want to pay for results – not for time spent.

This shift puts pressure on both margins and talent models. If half the work on a project can be automated in seconds, firms can no longer simply fill timesheets with junior hours to drive revenue. That is leading some providers to experiment with subscription and value-based billing tied to outcomes, sharing more risk and reward with clients. At the organizational level, the high cost of AI investment is challenging the partnership model: firm leaders accustomed to steady profit pools are wary of funding massive tech overhauls that may not pay off immediately ([3]). In response, a few have even taken on outside investors – for instance, in 2024 Grant Thornton sold a stake to private equity to raise capital for technology upgrades ([4]). The era of easy leverage is ending, and efficiency gains are being passed back to clients.

All of this raises an existential question: what truly is the value of expert human judgment when machines handle so much analysis? The answer may lie in exactly those areas where AI falls short. Leading firms now emphasize the importance of combining artificial intelligence with uniquely human insight. KPMG’s UK advisory head Lisa Fernihough, for example, says the firm’s edge comes from 'combining cutting-edge technology with deep sector expertise' ([5]) – essentially, pairing AI’s capabilities with seasoned human judgment. In the same vein, PwC leaders argue their highest value to clients is understanding how industries work in practice and navigating complexities that purely digital solutions can overlook ([6]).

However, even this human advantage is being tested. Nothing prevents AI companies from hiring top consultants and industry experts themselves – effectively internalizing the very human expertise that established firms bank on. One recruiter pointed out that firms like Anthropic or OpenAI can simply bring in leading consulting partners and 'deliver that service far more cheaply because they haven’t got to pay all the people in the teams that still exist within these larger consulting firms' ([7]). In short, professional advisors must focus on the irreplaceable elements of their craft – strategic creativity, ethical oversight, and trusted client relationships – to remain indispensable. The coming era belongs to those who can harness AI’s power while still providing the wisdom and judgment that machines lack.