CKO-026Hallucinations, Bias & FairnessStrong evidence
What is algorithmic bias?
Algorithmic bias occurs when an AI system produces systematically distorted, unfair or inaccurate outputs due to biases in data, design or implementation.
In more detail
AI systems learn patterns from data and processes created by humans. If training data, labelling decisions or system design contain biases, these can become embedded in model behaviour. In evidence synthesis, algorithmic bias may affect retrieval, screening, extraction or summarisation, potentially leading to systematic distortions in findings.
Why it matters
Bias can affect which evidence is found, included or emphasised.
Decision rule
Evaluate bias risks before deploying AI systems.
Common misconception
“AI is inherently objective.”
At a glance
- Evidence strength
- Strong
Related concepts
Key takeaway
AI can inherit and amplify human biases.