RAND researchers cite faulty data policies among reasons for AI failures

Researchers at the RAND Corp. say developers of artificial intelligence often lack the necessary data to train AI models, in a new report that examines the “root causes” of why AI projects often fail, while offering key steps for industry to avoid such pitfalls.

“Too often, trained AI models are deployed that have been optimized for the wrong metrics or do not fit into the overall business workflow and context,” says the Aug. 13 report titled “The Root Causes...