Data bias in AI

Artificial intelligence can sometimes be biased to certain shapes or colours. When such AI systems are applied to situations that involve people, then this bias can manifest itself as bias against skin colour or gender. This lesson explores bias in AI, where it comes from and what can be done to prevent it.

Additional details

Year band(s)	5-6, 7-8
Content type	Lesson ideas
Format	Web page
Core and overarching concepts	Digital systems, Data representation, Specification (decomposing problems), Impact and interactions
Australian Curriculum Digital Technologies code(s)	AC9TDI6K01 Investigate the main internal components of common digital systems and their function AC9TDI6K02 Examine how digital systems form networks to transmit data AC9TDI8K02 Investigate how data is transmitted and secured in wired and wireless networks including the internet AC9TDI6K03 Explain how digital systems represent all data using numbers AC9TDI8K04 Explain how and why digital systems represent integers in binary AC9TDI6P01 Define problems with given or co developed design criteria and by creating user stories AC9TDI8P04 Define and decompose real-world problems with design criteria and by creating user stories AC9TDI6P06 Evaluate existing and student solutions against the design criteria and user stories and their broader community impact AC9TDI8P10 Evaluate existing and student solutions against the design criteria, user stories and possible future impact
Technologies & Programming Languages	Artificial Intelligence
Keywords	Artificial Intelligence, AI, artificial, intelligence, teachable machine, smart phone, algorithms, problem solving, digital systems, Scratch, Lesson idea, Lesson plan, Digital Technologies Institute, data bias
Organisation	ESA
Copyright	Creative Commons Attribution 4.0, unless otherwise indicated.