Everyone talks about AI bias, how much should I actually worry about it?

More than most people do. Bias isn't some abstract academic problem, it shows up in who gets recommended for a loan, flagged for fraud, or denied access.

There are several potential risks of biases. For example, any decision by AI that ends up making unethical discrimination can land you on legal and compliance issues. Such biases and their fallout can also damage your business reputation. So, you need to take biases seriously from early on.