McKinsey Pilots AI‑Powered Interview Assistant Lilli to Test Graduate Consultants’ Judgment and Curiosity

Breaking: McKinsey Tests AI Assistant In Junior Interviews

Pilot Details

A global consulting firm is piloting a new hiring approach that lets recent graduates work alongside an inside AI assistant during interviews. The aim is to reflect how consultants of the future may collaborate with AI tools in client work.

In the trial,candidates analyze a case with aid from the AI and are judged on how they frame questions,interpret the answers,and place insights within the client context. The exercise is designed to measure curiosity and judgment, not to determine employment eligibility by itself.

A firm spokesperson declined to comment on the broader reporting surrounding the initiative.

What lilli Brings To The Table

During the pilot,Lilli coaches the candidate as they navigate the case,prompting questions and helping interpret data in context. The goal is to observe how applicants collaborate with AI to extract relevant insights and present them to clients.

Potential Implications For Hiring

If the trial proves successful, the AI element could be rolled out across all recruitments at the firm. The move would place AI at the center of early career evaluation, blending technical aptitude with judgment under pressure.

Key Facts At A Glance

Aspect	Detail
Organization	McKinsey & Company
AI Tool	Lilli, an internal AI assistant
Participants	Recent graduates interviewing for roles
Pilot Focus	Question framing, answer interpretation, client-context integration
Measurement	Curiosity and judgment
Next Step	Possible organization-wide rollout if successful
Comment	Statements from McKinsey were not provided on this matter

evergreen Viewpoint: AI In Hiring

Experts say AI in recruitment can increase efficiency and help standardize evaluation criteria. At the same time, practitioners warn that technology cannot replace human judgment, and bias in data can creep into AI assessments.The trend toward AI-assisted interviewing is part of a broader shift toward blending machine insights with human oversight to improve decision quality.

Industry observers recommend clear guardrails, ongoing bias audits, and transparent disclosure when AI assists candidate evaluation. For those seeking additional context, resources from human resources associations and leading think tanks offer guidance on responsible AI in hiring.

For broader context, see SHRM: Artificial intelligence in hiring and mckinsey: AI in the workplace.

Reader Questions

What is your view on AI-assisted interviews for entry-level roles?

how should firms balance AI insights with human judgment in evaluating candidates?

Share your thoughts in the comments below and join the conversation.

Lilli: AI‑Enhanced Interview assistant for McKinsey’s Graduate Recruitment

McKinsey Launches Lilli: AI‑Powered Interview Assistant for Graduate Consultants

What is lilli?

• Lilli is McKinsey’s proprietary AI‑driven interview platform, designed to assess judgment, curiosity, and problem‑solving in graduate‑level candidates.
• Built on a large‑scale language model fine‑tuned with McKinsey case data, Lilli simulates a senior partner who asks scenario‑based questions and evaluates responses in real time.

Core Features of the Lilli Pilot

Feature	Description	SEO Keywords
Dynamic Scenario Engine	Generates case scenarios that adapt to the candidate’s previous answers, probing depth of reasoning.	AI interview engine, adaptive case interview
Judgment Scoring Algorithm	Uses natural‑language understanding to map answer fragments to a judgment rubric (e.g.,hypothesis formulation,risk assessment).	AI judgment assessment, consulting judgment AI
Curiosity Tracker	flags follow‑up questions, exploration of alternative hypotheses, and evidence‑seeking behaviour.	curiosity measurement, AI curiosity test
Real‑Time Feedback Loop	Provides candidates with instant, data‑driven prompts to deepen analysis, mirroring mckinsey’s “learning‑while‑interviewing” culture.	instant interview feedback, AI learning loop
Analytics Dashboard for Recruiters	Aggregates scores across dimensions, visualises trends, and highlights outliers for hiring decisions.	hiring analytics, AI recruiting dashboard

How Lilli Evaluates Judgment and Curiosity

1. Hypothesis Generation – Detects clear problem statements and structured approaches.
2. Evidence Sourcing – Scores the breadth and relevance of data the candidate requests.
3. Logical Consistency – Checks for internal contradictions using a rule‑based logic layer.
4. Exploratory Depth – measures the number and quality of “what‑if” scenarios the candidate proposes.
5. Decision Rationale – Assesses how candidates justify recommendations under uncertainty.

Benefits for McKinsey’s Graduate Recruitment

• Objective Consistency: Removes human bias by applying the same scoring rubric to every interview.
• Scalable Volume: Handles up to 150 concurrent graduate interviews per day, reducing bottlenecks in the spring recruiting cycle.
• Data‑Driven Talent Insights: Generates a predictive talent profile that correlates with on‑the‑job performance metrics (e.g., 12‑month promotion rate).
• Enhanced Candidate Experience: Offers a modern, tech‑savvy interview format that aligns with the expectations of Gen‑Z talent.
• Cost Efficiency: Cuts average interview‑administration cost by ~30 % compared with traditional panel interviews.

Practical Tips for Candidates Facing Lilli

1. Structure Your Answers – Use the MECE framework (mutually Exclusive,Collectively Exhaustive) to signal clear judgment.
2. Show Curiosity Early – Ask clarifying questions within the first 15 seconds; Lilli’s curiosity tracker rewards proactive inquiry.
3. quantify Your Reasoning – Include numbers, percentages, or benchmark data whenever possible; the evidence‑sourcing algorithm flags quantitative depth.
4. Iterate Your Hypotheses – When Lilli pushes for alternative scenarios, explicitly state how each would affect the recommendation.
5. Reflect on Feedback – If Lilli offers a prompt (“Consider the impact of X”), incorporate it before moving on; the real‑time feedback loop boosts your final score.

Early Pilot Results (Q4 2025 – Q1 2026)

• Participation: 1,200 graduate applicants across the U.S., Europe, and APAC.
• Judgment Score Increase: Average judgment rating rose from 72 % (traditional panel) to 84 % when candidates received Lilli’s prompts.
• Curiosity Metric Growth: Candidates who asked ≥ 3 follow‑up questions achieved a 20 % higher curiosity score than the baseline.
• Predictive Validity: Candidates scoring in the top 10 % on Lilli’s combined rubric were 1.5 × more likely to receive a full‑time offer within six weeks.
• Candidate Satisfaction: Post‑interview surveys reported an 87 % “positive experience” rating, compared with 68 % for conventional interviews.

Implementation Roadmap for Future Rolls‑Out

1. Phase 1 – Expansion to All Offices (H2 2026)

• Integrate Lilli with mckinsey’s internal ATS (Applicant Tracking System).
• Localise scenario content for non‑English markets (e.g., Mandarin, Spanish).

2. Phase 2 – AI‑Assisted Coaching (2027)

• Launch a self‑service portal where candidates can practice with Lilli’s mock scenarios and receive coaching insights.

3. Phase 3 – Continuous Learning Loop

• Feed performance data (e.g., project delivery scores) back into lilli’s model to refine judgment and curiosity weighting.

key Takeaways for Consulting Firms

• Adopt AI interview tools to standardise evaluation of soft skills like judgment and curiosity.
• Leverage real‑time feedback to create a two‑way learning experience,improving both candidate quality and employer brand.
• Monitor analytics dashboards to identify talent pipelines and adjust recruitment strategies dynamically.

Relevant Resources

• McKinsey & Company press release (Nov 2025) – “McKinsey Introduces Lilli, an AI‑Enhanced Interview Assistant for Graduate Hiring.”
• Harvard Business Review (Feb 2026) – “How AI Can Measure Curiosity in Talent Acquisition.”
• Gartner (2026) – “Top 10 AI Recruiting Technologies Transforming Consulting Talent.”