Team Management
Your New Teammate Has No Desk—and May Still Carry Yesterday’s Bias
A human-plus-machine workforce changes hiring, learning, people analytics, and the manager’s duty of care.
Manage human–AI work as a socio-technical system. Define accountable owners, audit data and outcomes, preserve meaningful human judgment, and give employees a voice in redesign.
An August 5 AI4 panel featuring workforce leaders from JPMorgan Chase, Pearson, Maybank, eHealth, and Humaans explored AI-based skills analysis, personalized learning, surveillance, bias, and fairness. The agenda reflects a practical shift: AI is not only a tool an employee opens; agents may carry out work across several systems. Yet the historical data guiding them can contain unequal opportunity, inconsistent ratings, or biased decisions. Automation can scale a good process. It can also give an old problem excellent response time.
Assign a business owner for every agent, plus technical, security, and risk support. Define the tasks it may perform, data it may access, approvals it needs, and conditions that stop it. Keep consequential employment decisions—hiring, promotion, discipline, termination, pay—with accountable human review. Give reviewers enough evidence and authority to disagree. Tell employees when AI affects work allocation or evaluation and provide a challenge route. A manager cannot outsource responsibility by saying, ‘The model decided.’ Models do not attend difficult appeal meetings for a reason.
Include employees in workflow mapping and identify where AI removes drudgery versus where it removes learning, discretion, or human connection. Test outcomes across demographic and role groups, not only overall accuracy. Limit surveillance to a defined, necessary purpose and set retention boundaries. Measure workload shifts because automation often creates new review and exception work. The strongest human–AI system makes accountability clearer, not blurrier. People should know what the agent did, why a person approved it, and where to go when the result feels wrong.