CNBC reported on 12 November 2026 that AI is putting pressure on consulting’s time-based fees and graduate recruitment pipeline. The issue reaches the core of the traditional model, where large teams of junior staff have done research, analysis and presentation work while firms charge clients by time and staffing.
Svilena Tzekova, KPMG’s London-based global head of consulting, said the industry is changing because knowledge that once formed the basis of consulting work is becoming more accessible. Clients can now reach a level of advice much faster than before, she said, raising their expectations of what consultants should provide.
Advice and pricing
Tzekova said KPMG had made substantial revenue from providing advice, but that part of the business is now under pressure. “We were making quite a lot of money by providing advice. I think that part of our business is literally nonexistent anymore,” she said.
The pricing model has not yet caught up. Clients increasingly ask to pay for measurable outcomes, but procurement teams still compare firms by the number of people assigned to a project and their hourly rates. That creates a mismatch between the way consulting work is being sold and the way many buyers still evaluate proposals.
AI also adds costs that are not captured by a simple day-rate model. Tzekova pointed to computing power and AI provider charges based on how much text models process, alongside the cost of employing people. She said firms are still working out how to build those inputs into a commercially workable outcome-based price.
Fixed deliverables and implementation work
Smaller AI-native consultancies are presenting the shift as an advantage against the Big Four firms Deloitte, PwC, EY and KPMG. Emiko Caerlewy-Smith, head of investor relations at London-based Elixirr, said her firm has never billed clients on a time-and-materials basis. In her view, clients should pay for fixed deliverables, outcomes and value rather than the time taken to produce the inputs.
Caerlewy-Smith also described a move from advising clients toward helping them build systems. In one recent project, Elixirr helped a global insurer build an AI-based motor insurance platform. She said the firm spent six months getting its data into a format that its AI models could fully use.
That data preparation point matters for larger firms as well as smaller ones. If internal knowledge, project history and client material are not in a usable form, AI tools cannot simply be added to an existing operating model. The larger the firm, the more complex that preparation may become.
Junior roles and judgement
The second pressure point is the graduate pipeline. Consulting has traditionally hired thousands of graduates each year and developed them through entry-level tasks such as research, data analysis and producing presentations. Those are among the tasks now being automated.
Claire Trachet, a former Deutsche Bank mergers and acquisitions banker and founder of London-based advisory firm Trachet, said AI helps junior employees do their jobs faster. She questioned how they build the experience needed to progress if they no longer do the foundational work themselves. She said professional judgement is a craft built through reputation and ownership of the work.
EY is seeing similar changes in audit. Paul Goodhew, EY’s London-based global assurance innovation and digital leader, said AI is changing work traditionally done by junior auditors, including checking figures against source documents. He said AI can now carry out the first preparation of that work for an audit professional.
EY is also developing new roles. Goodhew named data delivery professionals who prepare data for audit teams, product managers who help develop technology platforms, and specialists who understand clients’ technology strategies and risks.
KPMG’s Tzekova said the answer is not to stop graduate recruitment. Instead, junior consultants will need to take on more responsibility earlier as the shape of consulting firms changes. Caerlewy-Smith said traditional firms need to redeploy employees quickly because the economics of the staffing pyramid no longer work if junior tasks disappear without replacement.
For clients, the change affects how consulting work is bought and assessed. A proposal based on staffing levels may say less about value when AI can do parts of the research, analysis or first preparation. A proposal based on outcomes still needs clear assumptions about data readiness, human review, implementation responsibility and the new costs introduced by AI systems.
As consulting clients ask for outcomes while firms absorb AI compute costs and rebuild graduate routes, ECS’s CxO + Strategy topic deals with the operating-model choices behind those decisions through keynotes, peer panels and closed-door roundtables for decision-makers. The programme gives attendees sessions and peer discussions where pricing assumptions and training models can be compared with practitioner experience from AI adoption work. Attend with the commercial and staffing questions that are difficult to settle from a proposal document alone.
This article was written by the ECS editorial team with the assistance of Amal Agent, run.events' AI assistant, which supported the research, drafting and proofreading.