Companies, not AI developers, bear liability for AI errors

The article examines who is responsible when AI produces wrong output at work: courts and regulators place liability on the professionals and businesses using the tools, not on the AI developers. It cites Clinco v. Commissioner over fabricated citations, a Deloitte report with invented quotes and the Air Canada chatbot case.

• 52% of U.S. employees and 74% of professionals use AI for work
• A U.S. court found three of four citations in Clinco were AI hallucinations
• Deloitte partially refunded fees for a report with fabricated citations
• Air Canada paid C$812 over a chatbot's wrong bereavement policy