What Is Algorithmic Bias?
Understand how bias enters HR AI systems, explore six distinct sources with real-world examples, and trace bias through every stage of the machine learning pipeline.
Bias Source Identifier
Click any card to expand it and reveal real HR examples. Each source represents a distinct way bias can infiltrate your AI systems — most organisations face multiple sources simultaneously.
Bias in Action
Explore three real-world scenarios where bias manifests in HR AI systems. Click each scenario to see how bias operates, who it affects, and what makes it so difficult to detect.
How Bias Enters the ML Pipeline
Bias can enter at every stage of the machine learning pipeline. Click each stage to understand the specific risks and how they compound as data moves through the system.
Key insight: Algorithmic bias is not a bug — it is a mirror. AI systems trained on historical HR data will faithfully reproduce the patterns in that data, including decades of structural inequality. Recognising this is the first step: the goal is not to build a “neutral” algorithm, but to deliberately design systems that are fairer than the human processes they replace.