On a May morning this year, nearly two million people across the United Kingdom pointed their phones at hedgerows, garden feeders, and woodland canopies and asked an app to name what was singing. That single month of activity hints at the scale of what Cornell Lab of Ornithology now intends to harness. In an update that researchers hope will sharpen conservation tools worldwide, the free Merlin bird-identification app is to be linked directly to eBird — a citizen-science platform that already holds more than two billion bird observation records gathered since 2002. Every species a user "hears" through Merlin will, in future, be logged automatically to that global database, turning tens of millions of casual listeners into an involuntary but powerful monitoring network.

The technology behind Merlin converts birdsong into spectrograms — visual maps of sound frequency over time — and uses machine learning to match the distinctive shapes those patterns make to one of 2,066 identifiable species. The system has been refined to cover virtually all birds in the United States, Canada, and Europe, as well as the more widespread species across India, Central America, and South America. Since Cornell launched the sound-identification feature in 2021, the app has been downloaded more than 40 million times across 240 countries, a figure that rose from 33 million as recently as December 2024. The urgency behind the upgrade is not difficult to find.

Britain alone has lost more than 70 million birds over the past half-century, according to the British Trust for Ornithology — a collapse so severe that the soundscape audible in a typical garden in 1976 would be almost unrecognisable today.