Can AI Make Better People Decisions than Humans?

London Business School (institutional)
London Business School (institutional)May 14, 2026

Why It Matters

Adopting rigorous, data‑driven hiring practices can improve talent quality, reduce bias, and boost long‑term organizational performance, while unchecked AI reliance may reinforce flawed decision patterns.

Key Takeaways

  • Data abundance enables evidence‑based hiring over intuition in practice.
  • Cognitive biases distort early judgments, harming selection quality.
  • Structured interviews outperform unstructured “chemistry” chats in predicting performance.
  • Organizations lack post‑hire metrics, limiting feedback loops for improvement.
  • AI tools must distinguish correlation from causation to avoid selection bias.

Summary

The Think Ahead podcast episode examines whether artificial intelligence can out‑perform human intuition in people‑related decisions, focusing on hiring, promotion, and retention. Professors Isabelle Fernandez Mateo and Sergey Gurif, together with Hatti Sundaram of the data‑driven hiring platform Applied, discuss the shift from gut‑feel judgments to evidence‑based approaches powered by richer data and advanced machine‑learning models.

Key insights include the explosion of granular employee data—from application details to digital communication footprints—and the recognition that traditional intuition is riddled with cognitive biases such as first‑impression anchoring and similarity attraction. The guests argue that structured interview protocols, clear competency rubrics, and longitudinal tracking of hires provide more reliable predictors of future performance than unstructured “chemistry” conversations.

Illustrative quotes highlight the tension: Sundaram notes that “good data plus bad data equals noisy decisions,” while Fernandez Mateo warns that “correlation is not causation; selection bias can mislead even sophisticated AI models.” Real‑world examples, such as the “airplane test,” reveal how organizations attempt to blend experiential assessments with data but often fail to validate outcomes over six‑month or twelve‑month horizons.

The implications are clear: firms must invest in systematic post‑hire analytics, align incentives with long‑term performance metrics, and train managers to critically interrogate AI outputs. Without these safeguards, the promise of AI‑enhanced hiring risks perpetuating existing biases rather than eliminating them.

Original Description

Decisions about who to hire, promote and reward shape organisational performance, culture and fairness. For decades, we have heavily relied on intuition. As data and AI become more embedded in the workplace, leaders face a critical question: how should evidence and human judgement work together in people decisions?
Sergei Guriev, Professor of Economics and Dean of London Business School, is joined by Isabel Fernandez‑Mateo, Adecco Professor of Strategy and Entrepreneurship at London Business School, and Khyati Sundaram, Founder of Applied.
Together, they explore how data, analytics and artificial intelligence are reshaping people management across the employee lifecycle. Drawing on academic research and real‑world experience, the conversation examines how AI is changing behaviour for both employers and candidates, and why leadership judgement, context and trust still matter more than ever.
Watch the episode to understand:
- Why intuition‑driven people decisions often reinforce bias rather than predict performance
- How AI is transforming hiring and why it is creating new challenges for employers
- Where data improves fairness and outcomes, and where human judgement must still lead
- What responsible, evidence‑based people management looks like in practice
For more thought leadership and business insights from London Business School faculty and alumni, visit ⁠⁠⁠https://www.london.edu/think
Sign up to receive a curated selection of articles, podcasts and films direct to your inbox twice a month at ⁠⁠⁠https://www.london.edu/think?entry=true#subscribeScreen
Follow us on social media: ⁠⁠⁠

Comments

Want to join the conversation?

Loading comments...