In the last lesson, you explored how the way we think about AI shapes what we trust, question, or overlook. Now we’re going deeper—into the system itself.

AI bias isn’t about one rogue algorithm. It’s about a socio-technical system, shaped by human choices, imperfect data, and workplace context long before any output appears on screen. In this lesson, you’ll unpack how bias actually forms across that system and why it's less about catching "bad behavior" and more about spotting where influence lives.

Through real HR examples—resume screening, performance analysis, engagement tools—you’ll learn to see where bias enters, how it quietly compounds, and where your judgment becomes a critical circuit-breaker before it scales.

Learning objectives

By the end of this lesson, you’ll be able to:

• Explain why AI in HR should be seen as a socio-technical system, not a standalone tool
• Identify key system components where bias enters (e.g. data, metrics, labels, infrastructure, environment, and feedback loops)
• Recognize common bias types in HR—including historical, representation, measurement, and aggregation bias
• Distinguish between normal AI behavior and outputs that signal a real bias risk
• Use a systems lens to assess when an AI-supported recommendation needs a second look