Powerful joint statement on the use of AI from the FTC and other federal agencies.: https://www.ftc.gov/system/files/ftc_gov/pdf/EEOC-CRT-FTC-CFPB-AI-Joint-Statement%28final%29.pdf
tl;dr — `It wasn't me, it was the AI' is not a defense against discrimination or other illegal harm.
h/t @emilymbender
@ct_bergstrom @emilymbender That's going to open up a whole new cottage industry of "AI Breakers" who learn how to prompt AIs to break their limitations and make discriminatory statements just to sue... The lawyers are gonna be so rich.
@mzedp @ct_bergstrom @emilymbender Right, but a common use case will be replacing customer service phone banks with AI. If an AI offers one customer a discount, but not another, wouldn't that have strong potential to be discriminatory?
@ct_bergstrom @LouisIngenthron @emilymbender The problem is not "this AI generated a faulty output", the problem is acting on that faulty output. So, I doubt this will be an issue.