ALT Screenshot of a new paper entitled "Early warning of trends in commercial wildlife trade through novel machine-learning analysis of patent filing" with a full abstract (summarised): 'Unsustainable wildlife trade imperils 1000s of species, but efforts to identify and reduce these threats are hampered by rapidly evolving commercial markets. Businesses trading wildlife innovate to remain competitive, and the patents they file to protect their innovations also provide an early-warning of market shifts. Here, we develop a novel machine-learning approach to analyse patent-filing trends from 1970-2020 related to 6 taxa that vary in trade legality, threat, & use: rhino, pangolin, bear, sturgeon, horseshoe crab, & caterpillar fungus. We found 27,308 patents, showing 130% per-year increases, compared to a background rate of 104%. Innovation led to diversification, including fertilizer using illegal rhino horn, and novel pangolin farming methods. Stricter regulation did not reduce patenting.