
Cornell Lab Silent on How It Will Verify AI Bird Identifications Entering eBird
Cornell Lab's Merlin app will feed 40 million users' machine-learning bird identifications directly into eBird, the world's largest ornithological database with 2 billion records. The Lab has not disclosed whether AI submissions will be flagged separately, quality-filtered, or pooled with human-verified data. This gap matters: eBird underpins continental conservation policy. Without metadata logging and provenance chains, AI submissions become indistinguishable from observer records, risking scientific weight on unvetted data.
Published