AI clones consulting workflows

On March 19, 2026 journalists documented a wave of browser‑based AI “agents” that produce structured case frameworks, benchmarking tables and slide‑ready recommendations previously associated with elite strategy teams. (b17news.com) Investigative reporting found that close to 150 former consultants from firms including McKinsey, BCG and Bain were contracted last year to label and teach models how to perform entry‑level consulting tasks. (bloomberg.com) Prompt engineers and consultants on social platforms have published explicit templates that privilege XML‑style structure and multi‑section “mega‑prompts” to force models into analyst‑like deliverables, with one unrolled thread demonstrating XML tag strategies for Anthropic’s Claude. (threadreaderapp.com) The same community member threads include copy‑paste “mega‑prompts” optimized for Gemini 3.0 Pro and other models, with stepwise input sections for hypothesis, data, calculations and slide outlines used for reproducible case outputs. (threadreaderapp.com) Open-access consulting prep libraries and university recruiting pages updated for the 2026 cycle — hosting case books, interview templates and model solutions — are being repurposed as training corpora and prompt testbeds for these AI workflows. (case-prep.com; managementconsulted.com) At the same time, major firms are integrating AI into hiring: McKinsey has piloted an AI interview stage using an internal chatbot called “Lilli,” signaling firms are evaluating candidates’ ability to use AI as part of consulting assessment processes. (vault.com) Industry founders and ex‑firm insiders argue this shift is accelerating boutique, AI‑powered consultancies that claim growth rates multiple times those of legacy firms, with recent startups raising early rounds (one cited $3.6M) to build consultancy‑style AI products. (thefinancestory.com)