Back to Headlines
Environment
Jun 16, 2026
Analyzed by GPT OSS 120B

AI Accelerates Fight Against Plant and Fungi Extinction, Kew Report Finds

AI Summary
A new Kew report shows that AI and massive digitisation of herbarium specimens are turning the tide in the race to save threatened plants and fungi. The data reveal rapid phenological shifts and open a genomic goldmine, but also flag energy use and bias risks.

The Royal Botanic Gardens, Kew report highlights how AI‑driven digitisation is becoming a game‑changer for identifying and protecting vital plant and fungal species before they disappear, while warning of energy costs and data‑bias challenges.

AI‑driven digitisation reshapes plant and fungi conservation

Scientists are now able to scan, analyse and share millions of specimens online, allowing rapid identification of new species, extraction of genetic data from century‑old fungi, and tracking of flowering‑time changes worldwide.

  • Prof Alexandre Antonelli (Executive Director of Science, Kew) says the effort “barely scratches the surface” of undiscovered biodiversity.
  • Digitisation of all 7.4 million Kew specimens has been completed, contributing to a global pool of 145 million digital records.
  • AI models trained on 8 million digitised specimens detected an average shift of 2.5 days per decade in flowering times.

Quantifying the data: scale, risk and climate signals

The report quantifies the magnitude of the extinction threat and the data‑driven insights emerging from AI analysis.

  • About 40% of the 70,000 assessed plant species are at risk of extinction.
  • 330,000 assessed species remain un‑analysed, and an estimated 100,000 species have yet to be named.
  • For fungi, 90% of the estimated 2 million species are still unknown, with less than 1% assessed for risk.

Why AI matters for the global biodiversity crisis

AI accelerates identification of hard‑to‑spot taxa such as sedges and peat mosses, often outperforming specialists, and unlocks genetic information from specimens up to 180 years old—turning historic collections into a “genomic goldmine”. The technology also democratises access to collections in biodiversity hotspots like Madagascar, where digitising 37,000 specimens revealed centuries‑long ecological data.

What the next decade could hold for AI‑enabled conservation

Future progress hinges on expanding digitised coverage (currently under 16% of global herbarium holdings), curbing the energy footprint of AI datacentres (which now consume about 6% of electricity in the UK and US), and addressing bias by diversifying source data. Partnerships between tech firms, environmental NGOs, and governments will be essential to fund and guide the next wave of AI‑powered biodiversity research.