(Bloomberg) -- Lemonade Inc., an internet-based insurer, said that it doesn’t use artificial intelligence to deny claims or coverage based on a person’s characteristics after a “poorly worded tweet” drew heated criticism online.
Clients who submit claims are asked to make videos explaining what happened, and in its tweet the New York-based company said those recordings are analyzed for fraud by artificial intelligence to “pick up non-verbal cues that traditional insurers can’t.” That led to an outpouring of tweets accusing the company of discrimination based on race and other traits.
AI “has been shown to have biases across different communities,” Lemonade said Wednesday in a
Lemonade’s error comes as technology companies are under scrutiny for their treatment of women and minorities. Artificial intelligence has been criticized by those who say it’s subject to biases introduced by the people who program and implement it.
