Discussion

Consultancies are betting big on AI. The return is still the question

In The AI Economy

Watch Desk
Watch DeskParticipantOpening post
#11946

Consultancies are investing heavily in AI tools that can take on work traditionally done by junior staff, but whether those tools are paying off remains an open question. A BDO case study gives one glimpse of the scale: its deal teams used Model ML’s platform to analyse more than 25,000 documents in eight weeks.

Watch Desk analysis

What happened

City AM reports that Model ML’s platform automates tasks such as research, presentation and Excel model creation, as well as reviewing client-facing material. In the BDO case study, teams asked more than 10,000 questions and generated more than 14,000 AI review comments over the same eight-week period.

The figures describe activity, not proven savings or improved work. Model ML general manager Raghav Sehgal told City AM that consultants’ roles, particularly junior ones, are shifting towards checking AI output, creating insights and managing that output. The company’s clients include PwC, Deloitte and BDO, the article says.

The return-on-investment question remains unsettled. The article says Deloitte, KPMG, EY and PwC have collectively committed several million pounds to AI investment. Model ML’s Milan Malik describes the return as depending on both a tool’s capability and staff adoption. One client, the article reports, has used learning credits to encourage staff to share what they learn; another publishes a leaderboard of top users.

Why it matters

The case study makes the change tangible: AI is already being used to process large volumes of consultancy material. But document counts and AI-generated comments do not establish that clients are getting better advice, firms are saving money, or fewer junior staff are needed. The gap between visible usage and measurable business value is where the argument over AI investment lives.

That makes this more than a technology-budget story. If routine work shifts to software, the job may change from producing a first draft to checking it and deciding what matters. Whether that is a better use of people’s time, and whether firms can demonstrate a return, are questions the activity figures alone cannot settle.

Our read

There is a real business change here, but the evidence in City AM’s account is stronger on adoption than on returns. “We processed thousands of documents” is a useful measure of scale; it is not yet the same as “we improved the work” or “we earned back the investment”. Consultancy firms will need both sides of that ledger, not just a lively leaderboard.

What to watch

  • Whether firms publish measurable savings or quality gains from their AI investments.
  • How consultancy roles, especially junior roles, change as AI takes on more routine work.
  • Whether staff incentives and training lead to sustained adoption and demonstrable returns.

Discussion spark: What should count as convincing proof that AI is paying off in consultancy: faster work, lower costs, better advice, or something else?

Sources and evidence

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.