The ethics of AI and machine learning
Forrester released new research today that covers how businesses can remain ethical when it comes to artificial intelligence (AI) and machine learning and what organizations must do to identify and prevent harmful data discrimination.
Key findings from the report include:
- AI will bring ethics into the spotlight. In the next three to five years, Forrester predicts that businesses will need to self-regulate, AI vendors will place the compliance onus firmly on buyers, and machine learning will further expose human bias.
- Biased models harm customers, brands, and businesses. Transgressions could damage a brand’s reputation, cause consumers to go to competitors, and potentially cost businesses up to 4% of annual revenue.
- Machine learning models are as good as the data used to train them. Not all types of bias are harmful, as people have different preferences, behaviors, and propensities; however, if an algorithm doesn’t appropriately reflect a population or one variable cancels another out, the end result could discriminate against potential customers and chance their purchase decision.
Click here for more information