How AI is helping keep endangered species offline and in the wild
AI models are now searching for Schreger lines, a defining feature of elephant ivory, to identify authentic elephant ivory products, such as this bangle confiscated by the US Fish and Wildlife Service. © WWF-US/Keith Arnold
As wildlife traffickers continue to exploit the internet to their advantage, technology is building stronger defenses to stop them.
Scaling detection rates and reporting
In 2024, company members of the Coalition to End Wildlife Trafficking Online reported blocking 24.1 million prohibited wildlife listings and suspected illicit sellers since 2018. Closing out 2025, that number has skyrocketed to 63.3 million, thanks in large part to new AI detection efforts by the companies such as Alibaba and Meta. While these increases reflect seven years of collaboration by the Coalition to elevate responses to illegal wildlife trade, the use of AI language-based detection is helping companies identify prohibited content at speed and scale, as well as better understand collected data.
Enhancing image detection technologies
Companies are also exploring the use of photo-based automation to surface violating content. eBay is integrating image detection technology into its toolbox to enforce its wildlife products policy. In 2023, eBay trained its first image detection model focused on a single wildlife product, which proved very successful at proactive detection of restricted content containing that product. Since then, eBay has developed two additional models, targeting different wildlife products. One of the models has resulted in hundreds of listing removals since going live, and another standalone model for an additional product is expected to go live soon. For more information, see eBay’s Transparency Report.
Baidu and IFAW launched the AI Guardian of Endangered Species 2.0, which can identify images of target species and products online to detect illegal wildlife trade. This upgraded version recognizes 34 species, including both wildlife products and live animals, with an average identification accuracy of 86%. The AI Guardian screened 430,000 suspicious images, accurately identified 40,000 images of targeted wildlife products, and facilitated the removal of 15,000 illegal online postings.
Increasing eyes across the web
While AI is advancing Coalition efforts, people power is still needed. The Wildlife Cyber Spotter Program serves as the Coalition's extra eyes across the web, harnessing the passion of conservation supporters to help find what AI misses. Citizen scientists are trained to detect and report protected wildlife sales online, either through volunteer programs or company employee engagement initiatives. For example, each quarter WWF-Singapore conducts training sessions for over 100 volunteers. To date, volunteers have identified tens of thousands of suspicious listings, along with hundreds of new evasion tactics and illicit search terms. This growing dataset is used to train an AI model that enhances reporting accuracy for law enforcement and online companies to take action. Ultimately, the goal is for the AI to become a fully automated system that can independently detect, compile, and report suspicious wildlife content at a scale to help combat this harmful trade.
How you can help
Maybe you’re not a trade, crime, or tech expert, but are a skilled scroller who cares about wildlife. They need you too! If you see a suspicious wildlife listing online, you should report it directly to the platform you’re on, or to the Coalition’s reporting page. Data is power. The more listings of prohibited wildlife content reported to online companies, the more data they have to train and improve their AI models to catch traffickers before transactions happen.
When the Coalition launched in 2018, many company enforcement teams weren’t yet collecting data on actions taken to remove content specifically for protected wildlife as it wasn’t (and in most countries, still isn’t) required by law. Now with AI interventions, spotting tiger stripes and elephant Schreger lines is quicker than the click to purchase one.
About the Coalition
The Coalition to End Wildlife Trafficking Online launched in 2018 with conservation convenors WWF, TRAFFIC, and IFAW. Now with 41 company members comprising more than 50 digital platforms, this collaboration unites the tech industry to standardize prohibited wildlife policies, train staff to detect prohibited wildlife, enhance automated detection filters, and educate and empower users to report suspicious listings. At the core of this effort is sharing learning and best practices to avoid duplication and prevent wildlife traffickers from shifting activities from one platform to the next. The Coalition complements additional approaches, such as those of law enforcement agencies. Get involved at www.endwildlifetraffickingonline.com