Cornell's AI Bird Data Heads to eBird Without Clear Quality Controls

Machine-learning identifications from Merlin, an app used by 40 million people, will flow directly into eBird—a database of 2 billion bird observations that guides conservation policy across North America. Cornell Lab has not said whether these AI submissions will be marked separately or flagged for verification. That silence creates risk: without tracking their origin, AI records become indistinguishable from human observations, potentially undermining the scientific credibility eBird depends on.