Why some Decisions Need a Human in the Room—Even when AI Gets It Right

Why It Matters

Automating purely cognitive decisions can boost efficiency, but neglecting the human relational component risks reputational damage and employee distress, especially in high‑stakes contexts like terminations.

Key Takeaways

• AI can replace cognitive tasks in many industries
• Human interpersonal communication remains essential for sensitive decisions
• ANZ's automated layoff email sparked backlash and reputational risk
• Leaders must delineate cognitive vs. relational aspects before automating
• Misapplied automation can increase employee distress and legal exposure

Pulse Analysis

The acceleration of artificial intelligence across corporate decision‑making is undeniable. From fraud detection to medical image analysis, AI now matches or exceeds human performance, prompting executives to ask whether a machine can do the job better. The ANZ Bank incident serves as a cautionary tale: while the algorithm correctly identified the need for workforce reduction, the delivery method ignored the human need for dignity and clarity, turning a routine HR action into a public relations crisis.

What separates a successful automation from a blunder is the recognition of two distinct labor dimensions: cognitive and interpersonal. Cognitive work—data crunching, pattern recognition, risk scoring—can be delegated to algorithms that operate faster and without fatigue. Interpersonal work, however, involves empathy, context, and the nuanced reading of emotional cues. When a layoff notice bypasses a face‑to‑face conversation, it strips away the relational layer, amplifying anxiety and eroding trust. Psychological research shows that transparent, humane communication mitigates the negative impact of job loss, underscoring why a human touch remains irreplaceable in certain scenarios.

For leaders, the path forward is pragmatic rather than ideological. Conduct a task‑level audit to map each decision point to its cognitive and relational components. Deploy AI where it demonstrably outperforms humans—such as underwriting or diagnostic screening—while reserving human judgment for moments that require empathy, negotiation, or ethical nuance. Establish governance frameworks that mandate human oversight for high‑impact outcomes, and embed feedback loops to monitor employee sentiment. By balancing efficiency gains with the essential human element, organizations can harness AI’s power without sacrificing the trust and morale that underpin long‑term performance.