A New Ally for Conservation: Can AI Help Stop Wildlife Trafficking?

A single shark fin in an x-ray.

Illegal wildlife trafficking continues to escalate as a global conservation crisis, threatening species across both land and sea. Detecting smuggled wildlife products has long challenged enforcement agencies, but emerging technology may offer a powerful new advantage. Researchers have developed an artificial intelligence system designed to expose trafficked marine wildlife concealed within luggage, making it increasingly difficult for smugglers to operate unnoticed. The study, led by Dr. Vanessa Pirotta of Macquarie University and published in Frontiers in Ocean Sustainability, combined artificial intelligence with existing airport CT scanning technology to detect commonly trafficked marine products, including shark fins, dried seahorses, and sea cucumbers. To determine whether AI could reliably recognize the unique shapes and features of these items, the research team assembled specimens, some sourced directly from wildlife trafficking seizures, providing realistic examples of what border authorities may encounter. Using CT scanners already deployed at airports worldwide, the team generated 298 scans from 68 individual samples. Each specimen was scanned repeatedly under different conditions: concealed beneath clothing, wrapped in metal materials, or hidden among everyday objects such as toys to mimic real smuggling tactics. To further strengthen the system, researchers applied a method called Threat Image Projection, digitally embedding images of marine wildlife products into scans of ordinary luggage. This dramatically expanded the training dataset and created lifelike screening scenarios for the AI model. By the end of the process, the algorithm had been trained on thousands of simulated baggage images before being tested against entirely new scans it had never previously encountered.

Shark fin samples in front of the AI scanner.

The results demonstrated strong potential for the technology. The AI successfully identified shark fins with 95% accuracy and dried seahorses with 96% accuracy. Sea cucumbers proved more difficult to detect, though the system still achieved an accuracy rate of 86%. At the same time, the study highlighted the limitations that remain. The algorithm produced false alarms in some instances, resulting in an overall false positive rate of 13%. In practical terms, this means certain bags would still require manual inspection even when no illegal wildlife products are present. Rather than replacing human expertise, the researchers envision AI as an additional layer of support in wildlife enforcement. Their goal is a collaborative approach where machine-assisted screening works alongside traditional identification methods. Human officers, detection dogs, and established inspection techniques would continue to play a critical role, with AI serving as a tool to strengthen and accelerate detection efforts.

The application of artificial intelligence in the fight against illegal wildlife trafficking represents a promising step forward for global conservation efforts. By testing the system’s ability to identify shark fins, dried seahorses, and sea cucumbers, this study demonstrated how emerging technologies could strengthen the detection of trafficked wildlife products—particularly in marine environments, where illegal trade remains far less documented than trafficking involving terrestrial species. At the same time, the research highlights the challenges that remain before such systems can be widely implemented. Many airports around the world still rely on conventional two-dimensional scanning systems, as advanced CT scanners remain costly and inaccessible in many regions. In addition, the study was based on a relatively limited number of physical specimens, meaning larger datasets and further validation will be necessary before AI-assisted screening can be adopted at a global scale. Even so, the findings offer a glimpse into the future of wildlife protection, where technological innovation could become a valuable ally in conservation. As Dr. Pirotta emphasized, AI is not intended to replace traditional identification methods. Instead, the most effective defense against international wildlife trafficking may lie in combining advanced machine detection with the experience of human officers, detection dogs, and established inspection techniques.