Ethical and social issues around data collection, including AI

Learning objectives

Explain the risks to personal privacy when a program collects and stores personal data, and describe how a programmer can protect it, for example by collecting only what is needed, asking permission and storing it securely

Describe how programs that track their users, such as location services, can both benefit and endanger personal security

Define algorithmic bias and recognize it in a described program, where the same group of users keeps getting worse results because of the way the program was built

Explain how the method used to collect a data set can build bias into any conclusion drawn from it, and why a programmer should examine that method before extrapolating from the data

Explain how an artificial intelligence or machine-learning system trained on unrepresentative data can reproduce the bias in that data, and suggest steps that reduce it, such as more representative data, regular testing for bias and open statements of known limitations