- Repeated searches of markets on the dark web have found negligible trace of illegal wildlife products.
- This news is troubling, conservationists say, because it suggests that traders are content to sell wildlife products on mainstream websites like eBay, where they rely on the sheer volume of transactions and lack of regulation to mask their activity.
- Regulating the wildlife trade on sites like eBay can be complex because the legality of sales is difficult to establish.
- Machine learning — developing computer systems capable of monitoring and policing online transactions — holds promise for enforcement on the surface web, but is currently hampered by online market operators’ failure to engage with the issue.
Beneath the readily available surface web lurks the “darknet” (or “dark web”) — a secretive hub for anonymous exchanges that often involve illicit goods like narcotics and child pornography. In a 2016 study published in Conservation Biology, scientists searched these dark corners of the internet for illegally traded wildlife products, such as rhino horn and elephant ivory. But they found almost nothing.
A year later the researchers repeated their search. Again, they reported little sign of the illegal wildlife trade on the darknet. For conservationists, the news was troubling.
“Criminals can trade on the surface web without facing charges for the most part,” said Julio Hernandez-Castro, a co-author of the paper and a cybersecurity expert at Kent University, U.K. “That’s why they don’t feel compelled to move to the darknet as other criminals, selling drugs or firearms, are forced to.”
The dark web
Most internet users only see the tip of the digital iceberg — sites that are indexed and readily accessible through search engines like Google or Bing. This tier of the World Wide Web, otherwise known as the “surface,” accounts for only five percent of the internet’s total depth. Beneath that hides the dark web.
It uses undocumented domain addresses to hide the identity and location of users and conceal communication between them. And it can only be accessed using servers like The Onion Router, or TOR, named for its method of granting anonymity by burying user information in layers of encrypted code. As the 2016 study notes, even if it were possible to search through 10,000 addresses on the dark web per second, it would take 3.8 trillion years to catalogue them all.
Black markets such as Silk Road and AlphaBay have bloomed in these conditions, allowing traders to buy and sell anything.
However, when Hernandez-Castro and his colleagues trawled through 9,852 items on sale in darknet markets, they found only one wildlife product: San Pedro cactus (Echinopsis pachanoi), a plant caught in the flood of drug traffic due to its hallucinogenic properties.
A report by Interpol this year also found “limited but clear evidence” of dark-web trading in wildlife products from endangered species, such as rhino horn, elephant ivory and tiger parts.
Yet overharvesting remains the second-largest cause of global species decline and extinction. The illegal wildlife trade, worth $19 billion to $26.5 billion per year to transnational organized crime, is the world’s fourth most profitable illegal trade, after drugs, human trafficking and counterfeiting.
Dark web vs. surface web
For Hernandez-Castro and his team, the conspicuous absence of wildlife trade on the dark web suggests that traders are content to hide in plain sight on the surface by using popular auction sites such as eBay. Since online trading is poorly monitored, the traders simply rely on the sheer volume of transactions to mask their activity.
“This is a very sad state of affairs … The scale of trade on the darknet is, for now, heavily correlated to the success in policing on the surface web,” said Hernandez-Castro. “And hence very, very limited.”
Even in the relative openness of the surface web markets, regulating the wildlife trade can be complex because the legality of sales is difficult to establish. Hernandez-Castro explains that species protected by the Convention on International Trade in Endangered Species (CITES) cannot be legally sold without a permit, for instance. There are also rules under the World Trade Organization (WTO) and national laws to consider that further complicate where each sale stands in legal terms.
“It isn’t just CITES, which is what most people discuss [the illegal wildlife trade] in relation to,” said David Roberts, a fellow co-author and researcher at Kent University. “CITES permits are essentially void if the trade is not in accordance with national law.”
Since the internet has streamlined cross-border trade, the legality of each transaction can become entangled in the respective laws of the seller’s and buyer’s countries. For enforcement agencies, this means laboriously sifting through individual transactions for evidence of anything that could break national or international laws. The product’s ultimate origin is also important to consider. Is the ivory in a traded antique from the tusk of an endangered elephant, or from the bones of a less threatened species?
“When it comes to online trade it is even more difficult as you only have a phone to identify if it is ivory, then it’s a question of whether it is pre- or post-CITES,” Roberts said.
Monitoring online wildlife trade
To distinguish between legal and illegal trade, Hernandez-Castro and Roberts have proposed an application of machine learning. This involves training computer programs to recognize patterns and intercept transactions which are likely to break restrictions.
In fact, the team managed to detect illegal elephant ivory with 93 percent accuracy on an antiques section of eBay using an automated system they developed. Despite the proven potential of their prototype, the pair have struggled to attract the necessary funding to develop it as a tool for enforcement.
“There is not enough interest and resources to police it properly on the side of marketplaces,” said Hernandez-Castro.
eBay’s animal and wildlife products policy forbids the listing of pets and most live animals for sale, with some exceptions including insects and molluscs used for food or bait. Anyone wishing to sell animal parts, pelt or skin is advised to contact the U.S. Fish & Wildlife Service and “follow applicable laws”, but the trade in any species included in CITES Appendix I (including elephants, tigers and rhinoceros) is prohibited.
“The fact that I have students that with only 10 minutes training can find elephant ivory easily just goes to show [eBay’s] systems aren’t as effective as they could be,” Hernandez-Castro argued.
Mike Carson, a senior manager at eBay, recently defended his company’s record on wildlife trafficking by highlighting their success in removing over 25,000 listings for illegal goods this year. Carson claimed eBay had begun to work closely with experts from the International Fund for Animal Welfare to train staff in detection techniques, while their executives penned a letter to the European Commission in support of a total ivory ban throughout the EU in July.
Although frustrated with the rate of change, Hernandez-Castro conceded that progress was at least being made.
“To be fair, when we did our initial study… elephant ivory made up seven percent of items we searched through, now it’s around the one to two percent mark.”
Online ennui may enable the illegal trade, but Roberts believes that conservationists have also failed in apprehending wildlife crime as it crosses the technological threshold.
“Currently the agenda is focused on boots-on-the-ground enforcement and demand reduction, which has, in my opinion, a poor evidence base [of success],” Roberts said. “Online trade fits into a nebulous space between enforcement and demand reduction, and therefore falls through the gaps when it comes to funding research and intervention.”
Entire species have similarly fallen through cracks because of regulation that is overly concerned with charismatic fauna like tigers and rhinos, Roberts argues.
From his background in botany, he cites the example of cycads in South Africa: plants which do not garner as much public attention and have quietly disappeared through the wildlife trade. “Cycads are prehistoric,” he said. “They are architectural plants, so people want them for gardens, golf courses, hotels … Cycads sell for thousands of pounds and as a result, three species are now extinct in the wild, seven have fewer than 100 individuals, and 25 are critically endangered. One species [of cycad] declined from 9,600 individuals to 390 in a matter of years. Now imagine the same decline in a charismatic species.”
Philip Muruthi, vice president of species protection at the African Wildlife Foundation (AWF), however, defended the fixation with tigers and rhinos among international NGOs tackling wildlife crime, insisting that “one has to be pragmatic and focus efforts.”
“They should expand efforts while still continuing to track the most trafficked and valued products, like pangolins, rhino horn, elephant ivory and valuable timber,” he said.
But Roberts argues that as long as popular species and photogenic solutions continue to receive the lion’s share of attention and resources, teaching computers to regulate the activities of online markets will remain a difficult sell.
“No funder seems to want to pick up this ball,” he said.
• Harrison, J. R., Roberts, D. L., & Hernandez‐Castro, J. (2016). Assessing the extent and nature of wildlife trade on the dark web. Conservation Biology, 30(4), 900-904.
• Hernandez-Castro, J., & Roberts, D. L. (2015). Automatic detection of potentially illegal online sales of elephant ivory via data mining. PeerJ Computer Science, 1, e10.
• INTERPOL. (2017). Illegal Wildlife Trade in the Darknet. INTERPOL Global Complex for Innovation.
• Roberts, D. L., & Hernandez-Castro, J. (2017). Bycatch and illegal wildlife trade on the dark web. Oryx, 51(3), 391-399.
Header Image: The pangolin – a relative of the anteater whose scales are used in traditional Chinese medicine – is the world’s most trafficked mammal. Photo by Rhett A. Butler/Mongabay