By Felix Soldner, Bennett Kleinberg, Shane D. Johnson

The popularity of online shopping is steadily increasing. At the same time, fake product reviews are published widely and have the potential to affect consumer purchasing behavior. In response, previous work has developed automated methods utilizing natural language processing approaches to detect fake product reviews. However, studies vary considerably in how well they succeed in detecting deceptive reviews, and the reasons for such differences are unclear. A contributing factor may be the multitude of strategies used to collect data, introducing potential confounds which affect detection performance. Two possible confounds are data-origin (i.e., the dataset is composed of more than one source) and product ownership (i.e., reviews written by individuals who own or do not own the reviewed product). In the present study, we investigate the effect of both confounds for fake review detection. Using an experimental design, we manipulate data-origin, product ownership, review polarity, and veracity. Supervised learning analysis suggests that review veracity (60.26–69.87%) is somewhat detectable but reviews additionally confounded with product-ownership (66.19–74.17%), or with data-origin (84.44–86.94%) are easier to classify. Review veracity is most easily classified if confounded with product-ownership and data-origin combined (87.78–88.12%). These findings are moderated by review polarity. Overall, our findings suggest that detection accuracy may have been overestimated in previous studies, provide possible explanations as to why, and indicate how future studies might be designed to provide less biased estimates of detection accuracy. 

PLoS ONE 17(12): 2022

By Sergi D. Bray , Shane D. Johnson and Bennett Kleinberg

Deepfakes’ are computationally created entities that falsely represent reality. They can take image, video, and audio modalities, and pose a threat to many areas of systems and societies, comprising a topic of interest to various aspects of cybersecurity and cybersafety. In 2020, a workshop consulting AI experts from academia, policing, government, the private sector, and state security agencies ranked deepfakes as the most serious AI threat. These experts noted that since fake material can propagate through many uncontrolled routes, changes in citizen behaviour may be the only effective defence. This study aims to assess human ability to identify image deepfakes of human faces (these being uncurated output from the StyleGAN2 algorithm as trained on the FFHQ dataset) from a pool of non-deepfake images (these being random selection of images from the FFHQ dataset), and to assess the effectiveness of some simple interventions intended to improve detection accuracy. Using an online survey, participants (N = 280) were randomly allocated to one of four groups: a control group, and three assistance interventions. Each participant was shown a sequence of 20 images randomly selected from a pool of 50 deepfake images of human faces and 50 images of real human faces. Participants were asked whether each image was AI-generated or not, to report their confidence, and to describe the reasoning behind each response. Overall detection accuracy was only just above chance and none of the interventions significantly improved this. Of equal concern was the fact that participants’ confidence in their answers was high and unrelated to accuracy. Assessing the results on a per-image basis reveals that participants consistently found certain images easy to label correctly and certain images difficult, but reported similarly high confidence regardless of the image. Thus, although participant accuracy was 62% overall, this accuracy across images ranged quite evenly between 85 and 30%, with an accuracy of below 50% for one in every five images. We interpret the findings as suggesting that there is a need for an urgent call to action to address this threat. 

Journal of Cybersecurity, 2023, 1–18 

By Michael J. Frith  , Kate J. Bowers  , Shane D. Johnson 

Introduction: In response to COVID-19, governments imposed various restrictions on movement and activities. According to the routine activity perspective, these should alter where crime occurs. For burglary, greater household occupancy should increase guardianship against residential burglaries, particularly during the day considering factors such as working from home. Conversely, there should be less eyes on the street to protect against non-residential burglaries. Methods: In this paper, we test these expectations using a spatio-temporal model with crime and Google Community Mobility data. Results: As expected, burglary declined during the pandemic and restrictions. Different types of burglary were, however, affected differently but largely consistent with theoretical expectation. Residential and attempted residential burglaries both decreased significantly. This was particularly the case during the day for completed residential burglaries. Moreover, while changes were coincident with the timing and relaxation of restrictions, they were better explained by fluctuations in household occupancy. However, while there were significant decreases in non-residential and attempted non-residential burglary, these did not appear to be related to changes to activity patterns, but rather the lockdown phase. Conclusions: From a theoretical perspective, the results generally provide further support for routine activity perspective. From a practical perspective, they suggest considerations for anticipating future burglary trends 

Journal of Criminal Justice, v. 82, 2022

By Shane D. Johnson and  Manja Nikolovska

Objectives Routine activity theory suggests that levels of crime are affected by peoples’ activity patterns. Here, we examine if, through their impact on people’s on- and off-line activities, COVID-19 restriction affected fraud committed on- and off-line during the pandemic. Our expectation was that levels of online offending would closely follow changes to mobility and online activity—with crime increasing as restrictions were imposed (and online activity increased) and declining as they were relaxed. For doorstep fraud, which has a different opportunity structure, our expectation was that the reverse would be true. Method COVID-19 restrictions systematically disrupted people’s activity patterns, creating quasi-experimental conditions well-suited to testing the effects of “interventions” on crime. We exploit those conditions using ARIMA time series models and UK data for online shopping fraud, hacking, doorstep fraud, online sales, and mobility to test hypotheses. Doorstep fraud is modelled as a non-equivalent dependent variable, allowing us to test whether findings were selective and in line with theoretical expectations. Results After controlling for other factors, levels of crime committed online were positively associated with monthly variation in online activities and negatively associated with monthly variation in mobility. In contrast, and as expected, monthly variation in doorstep fraud was positively associated with changes in mobility. Conclusions We find evidence consistent with routine activity theory, suggesting that disruptions to people’s daily activity patterns afect levels of crime committed both on- and off-line. The theoretical implications of the findings, and the need to develop a better evidence base about what works to reduce online crime, are discussed. 

Journal of Quantitative Criminology, 2022.

By Muhammad Amien Ibrahim, Samsul Arifin and Eko Setyo Purwanto

Social media moderation is a crucial component to establish healthy online communities and ensuring online safety from hate speech and offensive language. In many cases, hate speech may be targeted at specific gender which could be expressed in many different languages on social media platforms such as Indonesian Twitter. However, difficulties such as data scarcity and the imbalanced gender-based hate speech dataset in Indonesian tweets have slowed the development and implementation of automatic social media moderation. Obtaining more data to increase the number of samples may be costly in terms of resources required to gather and annotate the data. This study looks at the usage of data augmentation methods to increase the amount of textual dataset while keeping the quality of the augmented data. Three augmentation strategies are explored in this study: Random insertion, back translation, and a sequential combination of back translation and random insertion. Additionally, the study examines the preservation of the increased data labels. The performance result demonstrates that classification models trained with augmented data generated from random insertion strategy outperform the other approaches. In terms of label preservation, the three augmentation approaches have been shown to offer enough label preservation without compromising the meaning of the augmented data. The findings imply that by increasing the amount of the dataset while preserving the original label, data augmentation could be utilized to solve issues such as data scarcity and dataset imbalance.

United States, Journal Of Computer Science. 2023, 9pg

BY SMARRELLI, GABRIELA; WU, DONGYI; BAAGO-RASMUSSEN, LINE; HARES, SUSANNAH; NAKER, DIPAK

From the document: "This brief has been developed to support conversations on addressing violence in and through education. It focuses specifically on violence against children in and around schools, while fully recognising that different forms of violence are interconnected, and occur in multiple settings including homes, communities, and online. The brief provides an overview of the magnitude and effects of violence in and around schools and a review of evidence-based interventions aimed at eliminating school-related violence. The brief should not be viewed as an expansive summary. Readers interested in further details are recommended to visit the studies cited in the references."

CENTER FOR GLOBAL DEVELOPMENT; COALITION FOR GOOD SCHOOLS; SAFE TO LEARN

From the document: "Cyber threat actors are aware of--and deliberately target--single points of failure. A compromise or failure of a Water and Wastewater (WWS) Sector organization could cause cascading impacts throughout the Sector and other critical infrastructure sectors [hyperlink]. There are many aspects of the large and complex WWS Sector that pose challenges to raising cyber resilience sector wide: [1] Governance and regulation involve a mix of federal and state, local, tribal, and territorial authorities. [2] Cybersecurity maturity levels across the sector are disparate. [3] Often, WWS Sector utilities must prioritize limited resources toward the functionality of their water systems over cybersecurity. [4] Universal solutions to cyber challenges in a diverse, target-rich, and resource-poor environment are unfeasible. To provide meaningful cybersecurity support to the WWS Sector that can help with these challenges, CISA [Cybersecurity and Infrastructure Security Agency]--in conjunction with the Environmental Protection Agency (EPA) and the Federal Bureau of Investigation (FBI), as well as the federal government and WWS Sector partners listed in this guide's acknowledgement section--has created this joint Incident Response Guide (IRG) for the WWS Sector. The unique value of this joint IRG is that it provides WWS Sector owners and operators information about the federal roles, resources, and responsibilities for each stage of the cyber incident response (IR) lifecycle. Sector owners and operators can use this information to augment their respective IR plans and procedures. By empowering individual WWS Sector utilities, the authors of this guide--CISA, FBI, EPA, and the acknowledged federal government and WWS Sector partners--aim to drive improvements to cyber resilience and incident response across the WWS Sector."

From the document: "The healthcare sector is particularly vulnerable to cybersecurity risks and the stakes for patient care and safety are particularly high. Healthcare facilities are attractive targets for cyber criminals in light of their size, technological dependence, sensitive data, and unique vulnerability to disruptions. And cyber incidents in healthcare are on the rise. For instance, HHS tracks large data breaches through its Office for Civil Rights (OCR), whose data shows a 93% increase in large breaches reported from 2018 to 2022 (369 to 712), with a 278% increase in large breaches reported to OCR involving ransomware from 2018 to 2022. Cyber incidents affecting hospitals and health systems have led to extended care disruptions caused by multi-week outages; patient diversion to other facilities; and strain on acute care provisioning and capacity, causing cancelled medical appointments, non-rendered services, and delayed medical procedures (particularly elective procedures). More importantly, they put patients' safety at risk and impact local and surrounding communities that depend on the availability of the local emergency department, radiology unit, or cancer center for life-saving care. [...] This paper provides an overview of HHS' [U.S. Department of Health and Human Services'] proposed framework to help the healthcare sector address these cybersecurity threats and protect patients."

Editor(s): D. Asher Ghertner, Hudson McFann, Daniel M. Goldstein

Security is a defining characteristic of our age and the driving force behind the management of collective political, economic, and social life. Directed at safeguarding society against future peril, security is often thought of as the hard infrastructures and invisible technologies assumed to deliver it: walls, turnstiles, CCTV cameras, digital encryption, and the like. The contributors to Futureproof redirect this focus, showing how security is a sensory domain shaped by affect and image as much as rules and rationalities. They examine security as it is lived and felt in domains as varied as real estate listings, active-shooter drills, border crossings, landslide maps, gang graffiti, and museum exhibits to theorize how security regimes are expressed through aesthetic forms. Taking a global perspective with studies ranging from Jamaica to Jakarta and Colombia to the U.S.-Mexico border ;Futureproof expands our understanding of the security practices, infrastructures, and technologies that pervade everyday life.

Contributors: Victoria Bernal, Jon Horne Carter, Alexandra Demshock, Zaire Z. Dinzey-Flores, Didier Fassin, D. Asher Ghertner, Daniel M. Goldstein, Rachel Hall, Rivke Jaffe, Ieva Jusionyte, Catherine Lutz, Alejandra Leal Martínez, Hudson McFann, Limor Samimian-Darash, AbdouMaliq Simone, Austin Zeiderman

Durham, NC: Duke University Press, 2020. 312p.

By Nanzi Zheng, Karen M. Abram,  Leah J. Welty; et alDavid A

Importance  Youths, especially Black and Hispanic males, are disproportionately affected by firearm violence. Yet, no epidemiologic studies have examined the incidence rates of nonfatal firearm injury and firearm mortality in those who may be at greatest risk—youths who have been involved with the juvenile justice system.

Objectives  To examine nonfatal firearm injury and firearm mortality in youths involved with the juvenile justice system and to compare incidence rates of firearm mortality with the general population.

Design, Setting, and Participants  The Northwestern Juvenile Project is a 25-year prospective longitudinal cohort study of 1829 youths after juvenile detention in Chicago, Illinois. Youths were randomly sampled by strata (sex, race and ethnicity, age, and legal status [juvenile or adult court]) at intake from the Cook County Juvenile Temporary Detention Center. Participants were interviewed at baseline (November 1995 to June 1998) and reinterviewed as many as 13 times over 16 years, through February 2015. Official records on mortality were collected through December 2020. Data analysis was conducted from November 2018 to August 2022.

Main Outcomes and Measures  Participants self-reported nonfatal firearm injuries. Firearm deaths were identified from county and state records and collateral reports. Data on firearm deaths in the general population were obtained from the Illinois Department of Public Health. Population counts were obtained from the US census.

Results  The baseline sample of 1829 participants included 1172 (64.1%) males and 657 (35.9%) females; 1005 (54.9%) Black, 524 (28.6%) Hispanic, 296 (16.2%) non-Hispanic White, and 4 (0.2%) from other racial and ethnic groups (mean [SD] age, 14.9 [1.4] years). Sixteen years after detention, more than one-quarter of Black (156 of 575 [27.1%]) and Hispanic (103 of 387 [26.6%]) males had been injured or killed by firearms. Males had 13.6 (95% CI, 8.6-21.6) times the rate of firearm injury or mortality than females. Twenty-five years after the study began, 88 participants (4.8%) had been killed by a firearm. Compared with the Cook County general population, most demographic groups in the sample had significantly higher rates of firearm mortality (eg, rate ratio for males, 2.8; 95% CI, 2.0-3.9; for females: 6.5; 95% CI, 3.0-14.1; for Black males, 2.5; 95% CI, 1.7-3.7; for Hispanic males, 9.6; 95% CI, 6.2-15.0; for non-Hispanic White males, 23.0; 95% CI, 11.7-45.5).

Conclusions and Relevance  This is the first study to examine the incidence of nonfatal firearm injury and firearm mortality in youths who have been involved with the juvenile justice system. Reducing firearm injury and mortality in high-risk youths and young adults requires a multidisciplinary approach involving legal professionals, health care professionals, educators, street outreach workers, and public health researchers.

JAMA Netw Open. 2023;6(4):e238902. doi:10.1001/jamanetworkopen.2023.8902

By The Anti-Defamation League Center on Extremism

Every year, individuals with ties to different extreme causes and movements kill people in the United States; the ADL Center on Extremism (COE) tracks these murders. Extremists regularly commit murders in the service of their ideology, to further a group or gang they may belong to, or even while engaging in traditional, non-ideological criminal activities.

In 2023, domestic extremists killed at least 17 people in the U.S., in seven separate incidents. This represents a sharp decrease from the 27 extremist-related murders ADL has documented for 2022—which itself was a decrease from the 35 identified in 2021. It continues a trend of fewer extremist-related killings after a five-year span of 47-79 extremist-related murders per year (2015-2019). One reason for the trend is the decrease in recent years of extremist-related killings by domestic Islamist extremists and left-wing extremists.

The 2023 murder totals include two extremist-related shootings sprees, both by white supremacists, which together accounted for 11 of the 17 deaths. A third shooting spree, also by an apparent white supremacist, wounded several people but luckily did not result in fatalities.

All the extremist-related murders in 2023 were committed by right-wing extremists of various kinds, with 15 of the 17 killings involving perpetrators or accomplices with white supremacist connections. This is the second year in a row that right-wing extremists have been connected to all identified extremist-related killings.

Two of the incidents from 2023 involved women playing some role in the killing or its aftermath. This report includes a special section that examines the role played by women in deadly extremist violence in the United States by analyzing 50 incidents from the past 20 years in which women were involved in some fashion in extremist-related killings

New York: ADL, 2024. 36p.

By Julia Deeb-Swihart

Human trafficking is a serious crime that continues to plague the United States. With the rise of computing technologies, the internet has become one of the main mediums through which this crime is facilitated. Fortunately, these online activities leave traces which are invaluable to law enforcement agencies trying to stop human trafficking. However, identifying and intervening with these cases is still a challenging task. The sheer volume of online activity makes it difficult for law enforcement to efficiently identify any potential leads. To compound this issue, traffickers are constantly changing their techniques online to evade detection. Thus, there is a need for tools to efficiently sift through all this online data and narrow down the number of potential leads that a law enforcement agency can deal with. While some tools and prior research do exist for this purpose, none of these tools adequately address law enforcement user needs for information visualizations and spatiotemporal analysis. Thus to address these gaps, this thesis contributes an empirical study of technology and human trafficking. Through in-depth qualitative interviews, systemic literature analysis, and a user-centered design study, this research outlines the challenges and design considerations for developing sociotechnical tools for anti-trafficking efforts. This work further contributes to the greater understanding of the prosecution efforts within the anti-trafficking domain and concludes with the development of a visual analytics prototype that incorporates these design considerations.

Dissertation. Atlanta: Georgia Institute of Technology, 2022.

By Bridget Harris & Delaine Woodlock

This project explored the impact of technology on victim–survivors of intimate partner violence in regional, rural or remote areas who are socially or geographically isolated. Specifically, it considered the ways that perpetrators use technology to abuse and stalk women, and how technology is used by victim–survivors to seek information, support and safety. Interviews and focus groups with 13 women were conducted in regional, rural and remote Victoria, New South Wales and Queensland. The findings showed that perpetrators used technology to control and intimidate women and their children. While this impacted women and children’s lives in significant ways, causing fear and isolation, the use of technology was often not viewed as a serious form of abuse by justice agents. 

Australia, Institute of Criminology. 2022, 81pg

By Kristin M. Tennyson, Lindsey Jeralds, Julie Zibulsky

Congress requires courts to consider several factors when determining the appropriate sentence to be imposed in federal cases, among them the “history and characteristics of the defendant.” The sentencing guidelines also specifically authorize judges to consider an offender’s age when determining whether to depart from the federal sentencing guidelines. In this report, the Commission presents information on a relatively small number of offenders who were aged 50 or older at the time they were sentenced in the federal system. In particular, the report examines older federal offenders who were sentenced in fiscal year 2021 and the crimes they committed, then assesses whether age was given a special consideration at sentencing. This report specifically focuses on three issues that could impact the sentencing of older offenders: age and infirmity, life expectancy, and the risk of recidivism.

Washington, DC: United States Sentencing Commission, 2022. 68p.

By Courtney Semisch

This publication provides data on the broad variety of economic crime sentenced under §2B1.1.  The Commission undertook a project to systematically identify and classify the myriad of economic crimes sentenced under §2B1.1 using offenders' statutes of conviction and offense conduct. The Commission used this two-step methodology to assign the 6,068 offenders sentenced under §2B1.1 in fiscal year 2017 to one of 29 specific types of economic crime. This publication provides, for the first time, data from this new project as well as a brief description of the study's methodology.

Washington, DC: United States Sentencing Commission, 2019. 87p.

By Courtney Semisch

For the first time, this report provides in-depth criminal history information about federal economic crime offenders, combining the most recently available data from two United States Sentencing Commission projects. Key findings are that the application of guideline criminal history provisions differed among the different types of economic crime offenders. Also, The extent of prior convictions differed among the different types of economic crime offenders. Federal economic crime offenders did not “specialize” in economic crime. The severity of criminal history differed for offenders in the specific types of economic crime. Only about one-quarter of federal economic crime offenders with prior convictions were not assigned criminal history points under the guidelines.

Washington, DC: United States Sentencing Commission, 2020. 56p.

By Tracey Kyckelhahn, Tiffany Choi

This study expands on prior Commission research by examining the geographic mobility of federal offenders. For this report, mobility is defined as having convictions in multiple states, including the location of the conviction for the instant offense. This report adds to the existing literature on offender criminal history in two important ways. First, the report provides information on how mobile federal offenders are, as measured by the number of offenders with convictions in multiple states. Second, the report provides information on the proportion of offenders with convictions in states other than the state in which the offender was convicted for the instant offense. The report also examines the degree to which out-of-state convictions in offenders’ criminal histories contributed to their criminal history score and their Criminal History Category.

Washington, DC: United States Sentencing Commission, 2020.  24p.

By Harjeev Kour Sudan, Andy Man Yeung Tai, Jane Kim, and Reinhard Michael Krausz

Introduction: The Dark Web is a subsection of the Internet only accessible through specific search engines, making it impossible to trace users. Due to extensive anonymity, the drug trade on the Dark Web makes regulation complicated. We sought to uncover the scope of the online drug trade on the Dark Web and the impact it may have on the dynamics of global drug trafficking.

Methods: We conducted a literature review to elucidate the availability and distribution of drugs on the Dark Web based on data reported in existing literature (n = 14) between September 2012 and June 2019. We simultaneously collected data about substances and listings from Dark Web cryptomarkets (n = 13) active between August 2022 and January 2023. Data from the literature review and the Dark Web scrape were combined to draw trends in the chronological availability and distribution of drugs between 2012 and 2023.

Results: The data collected from 13 cryptomarkets between late 2022 and early 2023 showed a relative change in substance distribution compared to 2012–2019, with a decrease in prescription drugs (from >20% to <5%) and a doubling of opioid listings (from 5.5% to 9.25%), while no major changes were observed on average during 2012–2019 according to literature.

Conclusions: The Dark Web warrants more attention in the analysis of the global drug trade. Understanding Dark Web drug markets can inform targeted interventions and strategies to reduce drug-related harms, while ongoing research is necessary to anticipate and respond to future changes in the landscape of the illicit drug trade.

Drug Science, Policy and Law Volume 9 January-December 2023

By Scott W. Duxbury and Dana L. Haynie

Although economic sociology emphasizes the role of social networks for shaping economic action, little research has examined how network governance structures affect prices in the unregulated and high-risk social context of online criminal trade. We consider how over-embeddedness—a state of excessive interconnectedness among market actors—arises from endogenous trade relations to shape prices in illegal online markets with aggregate consequences for short-term gross illegal revenue. Drawing on transaction-level data on 16 847 illegal drug transactions over 14months of trade in a ‘darknet’ drug market, we assess how repeated exchanges and closure in buyer– vendor trade networks nonlinearly influence prices and short-term gross revenue from illegal drug trade. Using a series of panel models, we find that increases in closure and repeated exchange raise prices until a threshold is reached upon which prices and gross monthly revenue begin to decline as networks become over-embedded. Findings provide insight into the network determinants of prices and gross monthly revenue in illegal online drug trade and illustrate how network structure shapes prices in criminal markets, even in anonymous trade environments.

Socio-Economic Review, 2023, Vol. 21, No. 1, 25–50

By Ojasi Gopikrishna

What are the economics of the cyber drug trade that is rampant across the globe? What are the impacts that an illicit online drug trade can have on the economy? How does the economy take into consideration the unreported transactions on the dark side of the internet and what are the positive and negative implications of a widespread trade like this in these globalizing times? This research paper aims to analyze various data and information to find and conclude how the darknet drug trade has had significant economic impact on the international financial sector. This paper also elaborates on why the black market is highly prevalent in the current times and what are the ill-effects of Darknet, in particular, drug trade on the Darknet. The analysis of economic consequences of unreported dealings is also taken into consideration and elaborated on. The paper sheds light on the bleak and disturbing statistics of involvement of various countries in the online drug ring marketplace with countries like the United States, India and Turkey dealing heavily in the illegal Darknet drug markets. A statistical analysis to determine the economic calculations of the drug trade is depicted and conclusions about the economical structure of the dark net drug trade and markets and its impact through the times is elaborated upon. The ill-effects of the Darknet drug trade are highlighted and conclusions drawn in the paper depict the fluctuating condition of the online illegal drug trade market.

Unpublished paper, 2022.