factually
Index/Topics/Bias in AI systems

Bias in AI systems

The potential for AI systems to exhibit bias, particularly in the context of CSAM detection, and the need for mitigation strategies.

Fact-Checks

1 result
Feb 4, 2026

How do AI-based CSAM classifiers work and what are their accuracy and bias trade-offs?

classifiers combine perceptual hashing and machine-learning classifiers to find known illegal images and predict novel abuse imagery, but their strengths—scale and speed—come with measurable limits: h...