The big picture: A new study reveals most enterprises still rely on human oversight for AI-enabled processes, especially in HR and customer engagement, due to the complexities of human behavior.

Why it matters: Designing AI agents for people-centric functions like HR requires a nuanced approach beyond pure efficiency, as errors carry significant reputational and regulatory risks.

Between the lines:

• 73% of enterprises use humans-in-the-loop for AI-enabled employee training and coaching.
• Human-centric processes are less deterministic and more sensitive to perceived bias than 'thing'-centric processes.
• The cost of AI errors in HR (e.g., underpaying) can far outweigh the probability of error.

Staffing & HR impact: HR leaders must prioritize robust governance and risk management for AI tools, understanding that perceived fairness and compliance are paramount over raw predictive accuracy. This impacts talent acquisition, compensation, and employee relations strategies.

The bottom line: Effective AI in HR demands a 'think different' approach, prioritizing human-centric design and error consequence over mere efficiency gains.