Australia Tests AI to Stop Wildlife Smuggling

A seven‑month experiment conducted across Australia tested two artificial‑intelligence tools on real police seizures to determine whether computers could spot illegal animals in cargo and help officers act faster.

Key Findings

• Scope of Seizures
  116 creatures were seized, mainly reptiles and crustaceans from five taxonomic groups:
• Tiliqua (blue‑tongue lizards)
• Egernia
• Oedura
• Chelodina (turtles)

• Euastacus (crayfish)

• X‑ray Scanning
  Police scanned 48 packages with a specialized X‑ray machine, RTT110.
  The top AI model, AT.3, correctly flagged smuggled animals in 56% of the scans and produced 48 detailed 3‑D images that allowed experts to confirm hidden contents.

• Handheld Fluorescence Device
  33 blue‑tongue lizards were examined with a handheld X‑ray fluorescence device.
  Feeding the data into existing machine‑learning models enabled investigators to infer each lizard’s geographic origin.
  Results indicated that common blue‑tongue lizards were rarely caught in the wild, whereas shingleback lizards (Tiliqua rugosa) were more frequently captured.

• Impact on Policing
  Sharing AI results with law‑enforcement teams reduced the number of parcels slipping through postal services.
  The trial demonstrates that incorporating smart technology into investigative workflows can lower wildlife trafficking and support evidence‑based policing.

Takeaway

Smart AI tools, when integrated with advanced imaging technologies, can significantly improve the detection of illegal wildlife in cargo and empower law‑enforcement agencies to act more swiftly and effectively.