Study warns generative AI raises ML security and bias risks

A new study warns that cost-cutting use of generative AI in machine learning could heighten cyber-attack risks and expose sensitive data.