Consulting at a Crossroads – Evolution or Disruption in the Age of AI?

Foreword

Consulting is having a “both/and” moment. Both cyclical headwinds (slower IT spend, deferred transformations, cost-cutting) and secular tailwinds (AI, cloud, cyber, energy transition, supply-chain rewiring) are hitting at once. The result is uneven demand: clients are buying more help where technology, productivity and resilience are on the line, and less where traditional slide-driven diagnostics once dominated. AI is the catalyst and the accelerant. It compresses the old consulting workflow, research, analysis, synthesis, and delivery, while simultaneously expanding the surface area of problems clients want solved. In our earlier AI note, we argued that “augmented operators beat pure automation.” The same holds for consulting: firms that become AI-native operators (built on data, models, and software) will grow; firms that cling to a labour-arbitrage, report-centric model will shrink.

Why does this feel different from past cycles?

Every few years, the industry re-litigates the “end of consulting.” That didn’t happen after ERP commoditised process know-how, when Google put white papers at everyone’s fingertips, or when offshore delivery cut costs. But 2024–2025 is different in three ways.

First, clients are already using AI themselves. Procurement and operating leaders are piloting GenAI in contract drafting, RFP generation, spend analytics, supplier risk screening, and knowledge retrieval. When buyers show up with copilots and auto-summarised data rooms, the “information asymmetry” that once justified large discovery teams shrinks. Advisory value must move from finding things out to making the right things happen. Multiple practitioner studies from large firms show rapid GenAI experimentation in procurement and operations, from automated RFI/RFQ drafting to near-real-time spend dashboards. 

Second, AI is changing the supply side, how firms themselves work. One leading strategy house has publicly leaned into thousands of internal AI agents to automate research, analysis, and document production, while resizing junior pyramids and shifting a quarter of its book toward outcome-based engagements. Whether you see that as hype or harbinger, it signals a structural re-mix of leverage, margins and delivery models across the sector. 

Third, demand is bifurcating. Source Global Research expects global consulting to return to growth in 2025 after a disappointing 2024, but growth isn’t uniform, tech-enabled change, customer experience, cyber and resilience lead; broad cost-takeout decelerates. In the UK, for example, after contraction in 2024, analysts now see mid-single-digit growth as AI-linked work rebounds even as public sector demand tightens. That pattern, selective strength amid belt-tightening, is visible across developed markets.

What AI really changes in consulting (and what it doesn’t)

The old workflow is compressing. Task-by-task, GenAI reduces cycle time in scoping (“answer-first” hypotheses), fact packs (rapid literature scans and benchmark assembly), expert calls (auto-summarised transcripts), and draft deliverables (structured narrative generation). That doesn’t eliminate the need for consultants; it changes where they spend time. The junior-heavy pyramid that once funded “analysis-as-learning” is harder to defend when internal copilots can produce credible first drafts in minutes.

But the client's problem set is expanding. AI itself is a transformation topic (governance, data architecture, model risk, workforce enablement, use-case factories). It bleeds into core growth questions (pricing, personalisation, product design), operations (planning, sourcing, quality), finance (close automation, cash forecasting), and risk (cyber, compliance, AI safety). Surveys from the big firms show adoption curves steepening through 2024 and into 2025; leadership teams report positive ROI and expect to increase budgets even as they wrestle with trust, security and change management. 

Implementation is the centre of gravity. The winning firms are already acting less like think tanks and more like builders, with managed services, proprietary assets, alliances with hyperscalers and model providers, and skin-in-the-game pricing. Internal shifts at major strategy players, toward agents, outcome metrics, and implementation muscle are a tell. 

What won’t be automated: trust with the C-suite on ambiguous, politicised, high-stakes calls; organisational change at scale; cross-functional orchestration; navigating regulators and geopolitics; and designing operating models where technology is only half the story. AI narrows the gap on answers, not alignment.

The new buyer journey: from “tell me what to do” to “make it work”

Information abundance has flipped the first meeting. Ten years ago, a client might need a firm simply to “map the landscape.” Today, many arrive with AI-assisted landscape maps already in hand. External partners must earn their keep by (1) bringing privileged data or repeatable assets, (2) de-risking execution with accountable delivery, or (3) delivering credible capability transfer, not just recommendations. The strongest demand signals for 2025 line up with that pattern: customer growth, tech enablement, and security, domains where assets and implementation matter more than slide quality. 

Insourcing and global capability centres (GCCs) are rising. Enterprises keep building internal change engines, GCCs and “transformation offices” that blend engineering, analytics, and process talent. Providers still play a role, but as co-builders and accelerators rather than sole owners of the plan. 

The expert network + marketplace layer is maturing. On many diligence and “thin slice” questions, clients route around big-firm overhead and tap curated expert networks or on-demand talent platforms to move faster and cheaper. The expert network market itself rebounded in 2024 and is now a multi-billion dollar niche feeding both investors and corporates; talent marketplaces report steady enterprise uptake for “consulting 2.0” teams. 

Public sector buyers are also changing. Governments are writing bigger AI contracts and expecting measurable efficiency gains, often directly with tech vendors. That doesn’t displace consulting, but it reshuffles who leads and who follows on delivery. 

Economics under pressure: pricing, pyramids, and partnerships

Rate cards vs. outcomes. The more AI compresses traditional tasks, the harder it is to defend time-and-materials at historic rates for commoditizable work. You can already see two counter-moves: (1) outcome/outcomes-plus pricing where firms shoulder risk and share upside; and (2) asset-plus-services models that bundle proprietary tools with advisory and implementation. Some leading firms already say a sizable minority of their business is now outcome-based, a shift that aligns incentives and differentiates from marketplaces that only provide talent. 

The talent pyramid is flattening. If AI can do the first 30–50% of junior work, client teams get smaller and more spiky, fewer generalists, more “T-shaped” experts, and more software in the loop. That shows up in hiring pauses and selective resets across 2023–2024, now giving way to targeted 2025 hiring in AI, data, cyber, and cloud, particularly in the UK and parts of Europe. 

Capex and alliances matter. The firms investing heavily in AI tooling, data rights, and cloud partnerships will have margin leverage. PwC’s well-publicised US$1B GenAI program (and subsequent partnership moves) is emblematic: scaled upskilling + productized assets + hyperscaler alliances. Expect copycats, but the “bundle” is the story.

Macro still matters. After a softer 2024, multiple trackers expect mid-single-digit global growth in 2025, with stronger momentum in AI-adjacent work. But rate pressure persists in generic transformation and public sector work, and buyers will keep re-bidding for price and IP. 

How the major players are adapting

Strategy houses are leaning into AI-native delivery, internal agent ecosystems, thinner teams, and partnerships that blur advisory and build. The public narrative around one blue-chip firm, 12,000 internal agents, rightsizing headcount, 40% of revenue tied to AI-related work, and a push toward outcome contracts, signals the direction of travel. 

The Big Four are productizing transformation, standing up AI platforms, managed services, and industry-specific accelerators, and pushing massive training programs. PwC’s US roadmap (and a senior AI leader hire from Big Tech) illustrates how services firms are attracting tech DNA to lead internal and client-facing innovation. Deloitte’s survey work indicates enterprise adoption ramps into 2025; EY’s C-suite research reports widespread positive ROI and continued budget expansion.

Marketing-services adjacencies are converging with consulting. EY’s launch of a global Studio+ unit, blending design, data, CX, and tech, illustrates the growth side of the story: not just cost and controls, but end-to-end customer experience linked to revenue. Expect more cross-competition with digital agencies and cloud professional services.

Where AI bites deepest (and where it falls short)

Most at risk: report factories; generic market scans; undifferentiated benchmarking; broad “change readiness” assessments; slide-heavy storytelling. AI has already narrowed advantages here, and expert networks and marketplaces push price down further. 

Most defensible: regulated change (FS, healthcare, public); risk and cyber (where threat landscapes evolve weekly); large-scale tech modernization (architecture, data, MLOps); human capital (skill systems, productivity programs); and C-suite counsel on strategy and deals. These arenas demand judgment, operating experience, and accountability, areas where the right partners amplify internal teams rather than replace them. 

Most disruptive potential: the operating layer, where firms don’t just advise, they run something for you: a revenue engine, a model factory, a risk control, a procurement cockpit. As AI platforms mature, this becomes a recurring-revenue, software-plus-services business. The question is whether incumbents can change their DNA faster than product-native entrants can learn enterprise change.

A pragmatic playbook for consulting leaders

A pragmatic playbook for consulting leaders requires more than incremental adjustments; it calls for a fundamental rethinking of how firms deliver value in a world where AI is rapidly commoditizing information. The first priority is to build an AI-first delivery spine that moves well beyond isolated productivity pilots. Rather than relying on tools as an optional add-on, consulting firms must standardize the use of AI agents across research, synthesis, and quality assurance processes. The aim is not simply to produce “better slides,” but to radically accelerate time-to-impact. This means designing engagements so that data capture, reusable intellectual property, and insights can be built in from the very beginning. In practice, that shifts the consultant’s role from information broker to orchestrator of scalable, technology-driven solutions.

Another essential shift is in how consulting firms define their domain expertise. Instead of offering generalized advisory services across every sector, leading firms are beginning to own a handful of hard problems end-to-end. For example, in SaaS businesses, they may focus on net revenue retention, or in pharmaceuticals, on first-time-right quality processes. By packaging not only advisory know-how but also proprietary data rights, tailored playbooks, and embedded change management practices, consultants can credibly promise measurable outcomes rather than abstract recommendations. This integrated model, part software, part advisory, part operations, represents a significant evolution in what clients expect to pay for, and a more defensible position against AI-driven commoditization.

Finally, the structure of consulting teams themselves is under pressure to evolve. The traditional pyramid, built on large numbers of junior generalists leveraged by a few senior experts, is giving way to a leaner, more senior-biased model. Firms are finding that hiring operators who can “write and build”, professionals who understand both execution and communication, is more valuable than maintaining armies of PowerPoint producers. Every consultant will need to be fluent not just in business but also in AI prompt design and model evaluation. In this sense, the junior consultant of the future is no longer a slide-builder but a leverage point for senior specialists, capable of applying AI to extend impact rather than performing repetitive analysis.

Regional lens: Asia-Pacific (with a glance at Japan)

APAC is both the production floor and the demand frontier for AI-enabled consulting. India’s AI talent base and digital infrastructure position it for disproportionate influence in delivery and innovation, a point echoed by BCG’s leadership in recent remarks, useful context for firms building regional delivery hubs and for clients deciding where to anchor capability. Japan’s pattern, by contrast, is quality-focused adoption: careful governance, deep manufacturing use cases (yield, maintenance, energy), and public-sector modernisation that prizes security and explainability. Expect consulting-as-implementation to grow fastest where AI meets industrial productivity and citizen services, areas that reward trusted partners with local credibility and global playbooks. 

What we got right (and wrong) in our earlier AI note

In our AI piece a few weeks ago, we argued that winners would be those who combine human expertise with model-driven scale, not those who chase “pure automation.” The consulting market’s 2025 posture validates that view: the firms getting traction are the ones productizing their know-how, arming teams with agents, and pricing for outcomes, without pretending AI eliminates the messy human parts of change. Where we underweighted the risk was speed: AI compressed the bottom of the pyramid faster than many expected, forcing sharper choices on hiring and pricing in just 18 months.

Outlook: disruption, yes, displacement, no

Is AI an existential threat to consulting? Only to one kind of consulting. The slide-heavy, hours-billed, junior-leveraged version will keep losing share to products, platforms, and on-demand talent. But consulting as applied decision-making under uncertainty, backed by accountable execution, is in rude health. Several signals support a cautiously bullish view into 2026: enterprise AI budgets are expanding on the back of early ROI; public buyers are signing larger AI contracts with clearer delivery expectations; and independent trackers expect the market to re-accelerate in AI-adjacent service lines. The industry isn’t ending; it’s unbundling, and recombining, around software, data, and outcomes.

Bottom line

For consulting leaders, the mandate is stark: become an AI-native operator with differentiated assets and a delivery spine, or drift into a lower-margin, commoditised tier. For clients, the opportunity is equally clear: you can now buy results rather than slides if you structure the work that way. The crossroads isn’t a cliff; it’s a fork. Choose the path that builds capability, codifies IP, and shares risk. That’s where the next decade’s value will accrue.