Abstract
Introduction
The drug trade, marked by the unregulated availability and distribution of psychoactive substances, poses a significant global health risk (UNODC, 2022). Substance use is associated with a myriad of adverse health outcomes, such as psychiatric disorders, infectious diseases, and even death. In many cases, the illicit drug trade exacerbates the public health burden posed by substance use, as the substances available for use can be adulterated and unexpectedly potent. Markets for the illicit drug trade are often volatile and ever-changing, threatening systems of care that struggle to adapt proactively (Krausz et al., 2021). In attempts to curb the illicit drug trade, the last few decades have been marked by a global “war on drugs” (Cooper, 2015). However, the supply and demand of illicit drugs have persisted. Technological advancement has increasingly facilitated the illicit drug trade and added novel challenges to public health efforts to curb the negative consequences of unregulated substances (Petcu, 2017). Namely, the illicit drug trade progressively has shifted to online spaces such as the Dark Web, making it more difficult to regulate and contain (European Monitoring Centre for Drugs and Drug Addiction, 2016; Gupta et al., 2019; Orsolini et al., 2017).
The Dark Web is a subsection of the internet that has become a hub for illicit activities, such as drug trafficking, money laundering, and weapons trading. The Dark Web offers extensive layers of anonymity, making it appealing for individuals to engage in criminal behavior (Gupta et al., 2019). For example, information exchange over the Dark Web differs from the rest of the internet by employing robust encryption techniques for information exchange (Gupta et al., 2019). Additionally, the Dark Web is characterized by a network of websites that are not indexed by search engines, rendering it accessible exclusively through specialized software and browsers like The Onion Router (TOR) (Gupta et al., 2019).
The illicit drug trade on the Dark Web poses unique risks to public health and safety. Substances available on the Dark Web are unregulated and untraceable and, unlike traditional drug markets, are not geographically limited. Therefore, the Dark Web provides a global platform for drug trafficking. This has resulted in an increase in the availability and accessibility of illicit drugs, which may have contributed to drug epidemics around the world. The illicit drug trade on the Dark Web takes place on cryptomarkets, characterized by their use of cryptocurrency (e.g., Bitcoin) for payment (ElBahrawy et al., 2020). This form of payment is nearly untraceable and allows those selling and purchasing drugs on the Dark Web to evade law enforcement (Kumar and Rosenbach, 2019).
Silk Road, established in February 2011, is widely recognized as one of the earliest and most influential cryptomarkets (Soska and Christin, 2015). It introduced an innovative and efficient system for selling and purchasing illicit psychoactive substances, leading to the subsequent emergence of other online markets dedicated to drug trafficking (Orsolini et al., 2017; Tsuchiya and Hiramoto, 2021; Tzanetakis, 2018). Despite the takedown of Silk Road by the FBI in October 2013, its impact on the global drug trade remains significant, with subsequent markets experiencing substantial growth in revenues and transactions (Orsolini et al., 2017; Pergolizzi et al., 2017). Over the last decade, cryptomarkets have continued to evolve, offering a range of products from tobacco and marijuana to synthetic opioids and their analogs (Kruithof et al., 2016; Pergolizzi et al., 2017).
With reports of the drug trade increasingly moving online, especially onto the elusive Dark Web (Southwick, 2016), the characterization of these markets facilitating the move and trade is crucial to ensure that systems of care are able to facilitate timely prevention of and intervention for adverse outcomes of drug use (Krausz et al., 2021). Researchers have previously attempted to characterize various cryptomarkets, assessing and reporting variables such as drug availability, prices, sales, vendors, and user types. However, to our knowledge, there remains a limited understanding of how the availability and distribution of psychoactive substances have changed over time. Given the rapidly changing, dynamic nature of cryptomarkets and their potentially significant role in drug use patterns, further investigation into their evolution is imperative.
The objective of this paper is to characterize the Dark Web drug trade by uncovering the availability, distribution, and origins of illicit drugs on the Dark Web. Additionally, we seek to compare the availability of various substances now (2022–2023) and over the last decade (February 2012–October 2019). To achieve these goals, we collected data related to substance categories, relative distribution, and geographical origin from cryptomarkets on the Dark Web as well as data reported in the existing academic literature.
Methods
Literature review
We conducted a literature review to extract data pertaining to Dark Web drug markets reported in the existing literature. A search was conducted on four scientific databases: PubMed, Web of Science, PsycINFO, and Medline. Our search terms were grouped into three categories of terms relating to (1) the Dark Web, (2) psychoactive substances, and (3) markets and trades. A list of all search terms can be found in Supplementary Appendix A.
Our search yielded 5833 papers, and their titles and abstracts were screened by two authors independently to exclude papers that did not report on Dark Web drug markets. The full texts of the remaining 157 papers were then screened to further exclude papers not reporting on drug availability or those that reported parameters excluded in this paper (e.g., revenues, costs, and interviews).
The data reported in papers were included for descriptive analysis if they met the following inclusion criteria: (1) reported at least five of the following substance categories: cannabis/hash, stimulant, prescription, psychedelic, opioid, and ecstasy/MDMA, (2) reported the number or relative percentage of listings in each category of substances, and (3) reported the time frame during which data was collected.
Based on the inclusion criteria, a total of 13 papers were included for descriptive analysis and comparison in the present study. We manually extracted the number or relative percentage of listings in each of the included substance categories. Data reported for categories other than those specified in the inclusion criteria were summed in an “others” category of substances. Data from the literature review were organized chronologically to note trends in both the availability and distribution of illicit drugs on the Dark Web between September 2012 and June 2019.
Dark Web scrape
Simultaneous with the literature review, we performed manual data extraction from online drug markets pertaining to the distribution of the substances available. We followed the methods outlined in a previous study that characterized online weapons trafficking on cryptomarkets (Rhumorbarbe et al., 2018).
We first compiled a comprehensive inventory of Dark Web markets that engage in the trafficking of illicit drugs. Subsequently, we accessed these markets through a virtual private network (VPN) while creating individual user accounts for each website, with a total of 13 markets included in the study. We then extracted pertinent data from the Dark Web cryptomarkets manually. Specifically, we collected information on (1) the available categories of illicit substances, (2) the number of listings for each drug category, and (3) the respective geographic origins of these substances.
The extensiveness of the data we were able to extract was dependent on the availability of information on the various cryptomarkets. By aggregating listings from all the markets examined, we sought to gain a holistic perspective on the current availability and distribution of diverse illicit substances on the Dark Web. We also excluded markets dedicated to only one substance (e.g., cannabis only, psychedelic only), because we sought to examine the relative availability of various substances across the Dark Web.
Results
Literature review
The literature review yielded a total of 13 papers published between September 2012 and October 2019. One paper (Bhaskar et al., 2019) reported two distinct datasets, or snapshots of a given cryptomarket at a given time, which were included separately in the present study. All papers included figures or tables that reported on the direct numbers or relative percentages of listings for each selected category of substance (cannabis and hash, stimulant, prescription, psychedelic, opioid, ecstasy/MDMA, and others) on specified cryptomarkets.
Table 1 depicts the characteristics of the papers and datasets included in this study, the time period during which data were originally collected, the markets represented, and the total number of drug listings compiled in each dataset.
While 13 papers reported data collected from individual markets, one paper reported cumulative data collected from a total of 8 markets (Décary-Hétu et al., 2018). Silk Road and Silk Road 2.0 were overrepresented across the literature, being reported in 6/14 (42.9%) datasets (Bhaskar et al., 2019; Broséus et al., 2017; Dasgupta et al., 2013; Dolliver, 2015; Phelps and Watt, 2014). While data were reported for each year between 2012 and 2019, the year 2014 was overrepresented, being reported in 6/14 (42.9%) datasets (Bhaskar et al., 2019; Broséus et al., 2016; Dolliver, 2015; Norbutas, 2018; Nurmi et al., 2017; Rhumorbarbe et al., 2016).
The availability of substances on the Dark Web was generally consistent across cryptomarkets that sell drugs. All cryptomarkets reported in the literature offered categorical breakdowns of the types of substances on sale. Most cryptomarkets had similar categories of substances available, and some cryptomarkets further disaggregated data to identify specific substances within larger drug categories (Table 2).
Categories of substances and listings reported in the literature.
Cannabis and hash were the most abundantly available substances between 2012 and 2019, accounting for, on average, nearly one-quarter (22.7%) of all drug listings. Prescription drugs were found to be the second most abundant substance category, representing nearly one-fifth (20.4%) of all drug listings. Stimulants belonged to the third largest category (14.3%), greater than the category “others” (13.5%), which included all other drug types, including benzodiazepines and dissociatives, that were not categorized individually in the literature review. Ecstasy/MDMA (12.8%) and psychedelics (10.8%) were the fifth and sixth largest categories, respectively. The smallest drug category of its own reported in existing literature was opioids, accounting for 5.5% of all drugs listed for sale on the Dark Web. See Figure 1 for the average relative distributions of substances between 2012 and 2019.
Relative average distribution of substances between 2012 and 2019.
The chronological arrangement of datasets yielded from the literature showed that the relative distributions of all substances remained fairly consistent. Throughout various datasets over the years, cannabis and hash and prescription drugs generally remained the top two most available substances. Opioids persisted as the smallest individual category of substances listed. Figure 2 offers a visualization of the chronological substance distribution on the Dark Web.
Average chronological distribution of substances between 2012 and 2019.
In papers with a relatively small or negligible “others” category, the substances and their relative listing distributions reported can be seen as artifacts in Figure 2. Such was the case with Nurmi et al. (2017) who reported data collected between November 2014 to September 2015 from the Finnish version of Silk Road,
Dark Web scrape
A total of 13 marketplaces were scraped manually, yielding a total of 137,654 listings across various substance categories. A majority of cryptomarkets (10/13) sold miscellaneous items, including drugs. The characteristics and descriptive statistics of the data collected from each cryptomarket are listed in Table 3.
Characteristics and descriptive statistics of data collected from scrape.
The greatest proportion of listings belonged to the cannabis and hash category (29.5%). The second largest category of listings was stimulants (∼19.4%). Listings for psychedelics, opioids, and benzodiazepines were nearly equally abundant (9–9.5% each). ecstasy/MDMA followed closely behind, representing 8.8% of all listings. The smallest categories were prescription (4.3%), steroids (2.9%), and others (0.8%). See Figure 3 for a visualization of these distributions.
Distribution of substances listed for sale on the Dark Web between 2022 and 2023.
Comparing the Dark Web drug markets between 2012 and 2019 vs. now (2022–2023)
When combined, results from the literature review and the Dark Web scrape both indicate that the Dark Web is a dynamic landscape for the illicit drug trade. Results from novel data collected from cryptomarkets in this study between 2022 and 2023 show a changed distribution of various categories of substances in the years since 2019. Notably, the prescription category was found to be the third
Chronological distribution of substances on the Dark Web between 2012 and 2023.
Figure 5 depicts the categorical comparison of each major substance category between 2012–2019 and 2022–2023. Prescription drugs are more clearly seen as having decreased in their relative distribution and availability over the last decade. The categories of cannabis and hash, stimulants, opioids, and others have all increased in their relative distribution, whereas in addition to prescription, the substance categories ecstasy/MDMA and psychedelics seem to have decreased.
Historical (2012–2019) vs. present (2022–2023) relative distribution of categories of substances on the Dark Web.
Key findings
Descriptive analysis of the data collected through both the literature review and data scraping on the Dark Web indicated varying periods of volatility in the availability and distribution of psychoactive substances. Notably, no major changes were apparent in the distribution of substances, on average, between 2012 and 2019, as reported in the literature.
Contrarily, the novel data collected in this study from 13 cryptomarkets between late 2022 and early 2023 showed that there was a relative change in the distribution of substances from data reported in 2012 to 2019. Specifically, the category of prescription drugs decreased in proportion, becoming the second smallest proportion of listings, from over 20% between 2012 and 2019 to under 5% between 2022 and 2023. In addition, opioid listings seemingly doubled, growing from 5.5% to 9.25%.
Discussion
Contextual significance
Highly anonymized, elusive, and unregulated, the Dark Web is an ideal landscape for vendors selling illicit substances. The substances, in addition to the markets themselves, that are available on the Dark Web at any given time are difficult to predict. To offer insight into and characterize Dark Web markets, we conducted a study of the substances available on the Dark Web, both historically—between 2012 and 2019—and at present, between 2022 and 2023. Therefore, the present study offers, to our knowledge, a first snapshot in time of the categorical and chronological distribution and availability of various illicit psychoactive substances on Dark Web cryptomarkets.
These data are pertinent in the face of various overdose epidemics around the world and concurrent, ongoing changes in street drug markets, which pose challenges to the clinical system of care (Krausz et al., 2021). In North America, for example, market shifts have resulted in the increased availability of highly potent opioids like fentanyl, increasing the risk of overdose and death (Krausz et al., 2021). Unanticipated changes such as this can burden the healthcare system and lead to poor outcomes such as increased morbidity and mortality related to drug use by hindering the development of appropriate intervention and treatment (Krausz et al., 2021). In the absence of geographical limitations, fluctuations in Dark Web markets compared to street markets can pose an even greater and global threat to the healthcare systems designed to address substance-related issues.
Given their globalized nature, cryptomarkets may offer predictive value with regard to changes in street markets for illicit substances. It is possible that the substances available on the Dark Web reflect their supply and demand across street markets. For instance, researchers have found that cryptomarkets are frequently used for wholesale purposes, whereby drug vendors across street markets may purchase substances online and then sell to individual customers at the offline, street level (Aldridge and Décary-Hétu, 2016). Understanding Dark Web markets would offer insight into the supply chain of illicit substances and potentially inform predictions surrounding which substances are likely to dominate street markets (Broadhurst et al., 2021).
This study also has important implications for drug policy. The increasing availability of potent substances such as opioids on the Dark Web indicates that the current drug policies are ineffective. Criminalizing drug use facilitates opportunities for illicit drug markets to operate both on the streets and online, including highly elusive spaces like the Dark Web. As found in this study, despite efforts to curb the availability of illicit substances on the Dark Web, cryptomarkets list these substances for sale in abundance. The harms associated with the illicit drug trade may be mitigated by decriminalization of substance use and the availability of a regulated safe supply of psychoactive substances.
Limitations
Proactivity is challenged by the scarce literature on this topic, as well as methodological challenges due to the volatility of Dark Web markets. We found that there was scant data that reported consistently on the substances that are available on the Dark Web at any given time. However, through the extraction of listing-related data, including clippings of the Dark Web menu panels, we obtained sufficient data to represent each year at least once throughout 2012–2019.
The volatility of Dark Web markets also posed a challenge for data collection, which occurred from August 2022 to January 2023. Though 29 cryptomarkets were initially identified for data extraction, immediately prior to the period of data collection, nine markets had been shut down. Data from more numerous markets than included in this study may yield more robust results.
Future directions
A future systematic review of all literature reporting on substance listings would help to corroborate the findings of the present literature review. Additionally, a longitudinal study could be undertaken to depict the dynamic landscape of cryptomarkets and substances available on the Dark Web at various times in succession. Geographic specificity could be taken into consideration in a future study to identify the emerging patterns of substance distribution on the Dark Web in different regions. Furthermore, conducting qualitative research or interviews with Dark Web vendors and users could provide valuable insights into the motivations and strategies behind the distribution and consumption of illicit substances.
Another important future direction is to explore the relationship between changes in substance distribution on the Dark Web and their impact on public health outcomes, such as overdose rates and drug-related harms. Analyzing these correlations can help inform targeted interventions and harm-reduction strategies. Additionally, given the potential role of the Dark Web in exacerbating drug epidemics, further research is needed to investigate the effectiveness of interventions specifically targeted at reducing the supply and demand of illicit substances on online drug markets. This could involve collaborations with law enforcement agencies and policy-makers to develop evidence-based strategies that disrupt and dismantle these illicit networks.
Lastly, studying the impact of legislative and policy changes on the Dark Web drug trade can provide insights into the effectiveness of such measures and guide future policy decisions. Changes in drug regulation and classification, and corresponding changes in the availability and distribution of substances may allow for the evaluation of legislation on the drug trade. For example, future studies could explore whether the legalization of a psychoactive substance has regional effects on its online illicit trade. Future studies may also examine the opportunities uniquely offered by the availability of psychoactive drugs on the Dark Web, and not only the challenges posed by such a market as has been the topic of much research thus far.
Overall, future research should aim to deepen our understanding of the Dark Web drug trade, its evolving dynamics, and its impact on public health, while exploring innovative approaches to mitigate the harms associated with the illicit drug trade in the digital realm.
Conclusion
The findings corroborate existing reports in the literature on the volatility and instability of Dark Web markets and their listings. Further characterization of the illicit drug trade on the Dark Web is necessary to reveal the extent of its role in global drug crises, anticipate and monitor changes in local street markets, evaluate the impact of drug policy decisions, and develop effective strategies to reduce drug-related harms. Monitoring and research on online distribution and the impact on the dynamic of drug markets (e.g., using TOR) is a field of research to be developed.
Supplemental Material
sj-docx-1-dsp-10.1177_20503245231215668 - Supplemental material for Decrypting the cryptomarkets: Trends over a decade of the Dark Web drug trade
Supplemental material, sj-docx-1-dsp-10.1177_20503245231215668 for Decrypting the cryptomarkets: Trends over a decade of the Dark Web drug trade by Harjeev Kour Sudan, Andy Man Yeung Tai, Jane Kim and Reinhard Michael Krausz in Drug Science, Policy and Law
Footnotes
Declaration of conflicting interests
Funding
Supplemental material
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
