The darknet (or darkweb/any variation thereof) has an undeniable stigma. Some know the hidden sites to be a gateway to speaking freely. Others use darknet marketplaces to purchase drugs that are safer than those on the street. Another group may use the anonymity to share child pornography. In the words of Ross Ulbricht, “I learned… when you give people freedom, you don’t know what they’ll do with it.”
Researchers from Terbium Labs claim to have found evidence that disputes the majorly negative reputation the darknet has garnered.
“Anonymity does not mean criminality,” the study’s landing page displays. “In the industry’s first data-driven, fact-based research report, Terbium Labs analyzes what’s really taking place on the far corners of the Internet.”
Dr. Clare Gollnick and Emily Wilson, according to Engadget, claimed to be the first to conduct such a study. While the exact intention or scope of the claim remains unknown, Terbium Labs is far from the first entity to conduct a scan of onion links. Thanks to the well-known security and privacy researcher Sarah Jamie Lewis, we have OnionScan. And again, thanks to Lewis, we have a list of darknet papers and studies conducted throughout the last decade.
The Terbium Labs paper listed the the full methodology at the end of the paper, but the introduction holds “what you need to know to get started.” To start, Dr. Clare Gollnick and Emily Wilson used data pulled from 400 URLs. The URLs were pulled by an automated crawler over the course of a single day. Each URL, the paper noted, was used as an independent unit.
A team of analysts classified the contents of each URL. The categories were predefined were labeled with one of the following terms: Legal, Explicit, Drugs, Pharmaceuticals, Fraud, Multiple Categories (Illicit), Falsified Documents & Counterfeits, Exploitation, Hacking & Exploits, Weapons, Extremism, Weapons of Mass Destruction, Other Illicit Activity, Unknown/Site Down, Downloadable File.
Legal content made the majority of the 400 domains documented and it mirrored what could be found on the clearnet. According to the study, 6.8% of the legal content was porn. The rest consisted of nothing worthwhile. There were political blogs, graphic design firms, and even forums to discuss erectile dysfunction. The legal content appeared to be seemingly no different than content anywhere else.
After the legal category came everything else.
The majority of the content in the study is simply a description for each of the aforementioned categories. For example:
We defined Drugs as any non-pharmaceutical drug or substance bought or sold for recreational purposes. To provide a more detailed breakdown of the kinds of drugs available on the dark web, we separately classified any Pharmaceuticals available for sale as well. We include marijuana as a drug and not a pharmaceutical for the purposes of this study.
Pharmaceuticals include any kind of drug that a doctor might prescribe, excluding painkillers and their derivatives. For our classification, Pharmaceuticals include ADD/ADHD and anti-anxiety medications, even though these medications are often used recreationally… No prescriptions, unlimited refills, and no questions asked. Dark web pharmacies provide unfettered access to prescription medications, recalled over-the-counter drugs, and unregulated supplements.
Note that in this study, prescription drugs and street drugs were not categorized together.
The study found that the “drugs” category constituted close to 44.5% of the illegal content on the darknet.
Illegal pharmaceuticals only accounted for another 11.9%.
Both categories combined, the study found, made up the majority of the darknet content at around 56%.
For the most part, the remaining categories, save for “Multiple Categories (Illicit),” made little impact. “Weapons” and “Weapons of Mass Destruction,” among others, yielded no results in study. One category stood out to researchers: Exploitation.
Researchers discovered more content depicting the exploitation of children than content in several other categories. The exploitation category was almost as large as the fraud category.
The results of this study should not necessarily be treated as canon. Various scans over the years have had very different results. Most scans with accompanying data have been far more in-depth than this one. This doesn’t change the fact that legal content exists on the darknet. Similarly, this study’s inaccuracies and small sample size do not inherently disqualify other findings.
Sarah Jamie Lewis, on Twitter, pointed out some issues with these types of studies. One-third or more of darknet sites have a duplicate or clone, according to Lewis. In a study like the one from Terbium Labs, pulling any number of duplicates paints an inaccurate picture. Likewise, Lewis wrote that she had never seen a study that did anything other than http-only. Other factors like site ownership and a website’s weight need to be taken into account.