José Carlos Hernández-Gutiérrez

The phenomenon of organized crime in Mexico is not new. However, it is no less true that, for a few years now, violence caused by or related to criminal organizations has been registering higher levels than ever. This has caused, from different spheres, to wonder if indeed the problem facing the country can still be called organized crime. Can non-state armed actors in Mexico qualify as terrorists? Have they evolved into some form of insurgency? The author of these pages, after conducting a bibliographic review on the variants of terrorism and insurgency used by some authors to refer to the Mexican case, affirms that Mexican criminal organizations are not terrorists or insurgents, but rather profit-making organizations that make a tactical use of terrorism and / or insurgency to achieve economic benefits or the goals of their organizations.

2021, Los desafíos de la globalización: respuestas desde América Latina y la Unión Europea

By Andres Ham Gonzalez, Juanita Ruiz

We estimate the effects of drug-related violence on individual labor market outcomes in a transit country. Transit countries do not have enough market power to determine the global supply or demand of drugs yet must deal with the consequences of drug trafficking activities. We implement a Bartik-type instrumental variables strategy which assumes that violence in Honduran municipalities located along drug transport routes changes when coca production in Colombia grows or contracts. Our results show that drug-related violence has negative effects on extensive and intensive margin labor market outcomes for transit country workers and has greater effects on women than men.

IZA Working Paper 17126, Bonn: Institute of Labor Economics - IZA, 2024. 

By Sanna Bergvall, Núria Rodríguez-Planas:

Most empirical studies indicate that becoming a mother is an augmenting factor for the perpetration of intimate partner violence (IPV). Using rich population-wide hospital records data from Sweden, we conduct a stacked DiD analysis comparing the paths of women two years before and after the birth of their first child with same-age women who are several quarters older when giving birth to their first child and find that, in contrast to the consensus view, violence sharply decreases with pregnancy and motherhood. This decline has both a short-term and longer-term component, with the temporary decline in IPV covering most of the pregnancy until the child is 6 months old, mimicking a temporary decrease in hospital visits for alcohol abuse by the children's fathers. The more persistent decline is driven by women who leave the relationship after the birth of the child. Our evidence is not supportive of alternative mechanisms including suspicious hospitalizations, an overall reduction in hospital visits or selection in in seeking medical care, mothers’ added value as the main nurturer, or mothers’ drop in relative earnings within the household. Our findings suggest the need to push for public health awareness campaigns underscoring the risk of victimization associated with substance abuse and to also provide women with more support to identify and leave a violent relationship

Bonn: Institute of Labor Economics - IZA, 2024

By Michael McLaggan

The South African public transport sector is facing a critical crisis, marked by escalating violence and extortion within the taxi industry. This report delves into the dynamics behind these troubling incidents, shedding light on the severe impact they have on bus companies and the broader public transport ecosystem.

Since 2015, the Intercape Ferreira Mainliner bus company has experienced over 176 attacks, highlighting the pervasive nature of violence linked to the taxi industry. These attacks include stonings, shootings, and acts of intimidation, creating a climate of fear and insecurity for passengers and operators alike. The violence is not isolated, but part of a broader campaign of economic coercion and extortion aimed at controlling the transport market.

The taxi industry’s aggression towards long-distance bus companies stems from intense competition over routes and pricing. Taxi operators argue that bus companies, with their extensive networks and competitive pricing, are undermining their business. In response, some actors within the taxi industry resort to violent tactics to enforce their demands, including dictating where buses can stop and what prices they can charge. This extortion threatens economic freedom and disrupts the stability of the public transport system.

The report reiterates that the violence and extortion tactics employed by the taxi industry amount to organized crime. Despite numerous court orders and high-profile cases, the state’s response has been inadequate. Law enforcement and political bodies often show reluctance to intervene, partly due to complex relationships with the taxi industry. This lack of decisive action perpetuates the cycle of violence and undermines the rule of law.

To combat this crisis, the report recommends several critical actions:

Prosecution of Coordinated Attacks: Treating violent incidents as organized crime under the Prevention of Organized Crime Act (POCA).

Comprehensive Investigations: Multi-organizational task forces should investigate extortion in the public transport sector.

Engagement and Dialogue: Convening high-level panels involving all stakeholders to find peaceful resolutions to disputes.

The report underscores the urgent need for a coordinated, robust response to the violence and extortion plaguing South Africa’s public transport sector. By addressing these issues head-on, the state can restore safety, uphold economic freedom, and reinforce the rule of law, ensuring a secure and stable environment for all public transport users.

Geneva, SWIT: The Global Initiative Against Transnational Organized Crime 2024. 52p.

By Observatory of Illicit Markets and the Conflict in Ukraine.

This report assesses the organized crime risks associated with veterans in Ukraine. It is a complicated, sensitive subject: while the hot state of the conflict means that very few military personnel are being demobilized at present, there is also concern that discussing veterans in such a context may stigmatize them. But, as this report highlights, such risks cannot be ignored and preparations for demobilization now may help prevent negative outcomes in future. Our research identifies the following as key organized crime risks: the incidence of drug use among service personnel, the profusion of weapons in the country and the possibility of veterans being recruited into or forming organized crime groups or joining private security firms as muscle. More tangentially, a sense of disaffection among veterans – rooted in a perception that the state is not keeping its promises to provide individual support or reform society as a whole – may also drive a wedge between veterans and society, generating friction and increasing the risk of confrontation, perhaps with violence. Veterans policy in Ukraine is a fast-moving field. The Ministry for Veteran Affairs is spearheading the new veterans law – a crucial development to ensure that Ukraine’s legislation is fit for the new realities of a post-conflict period that will see a million or more veterans in society. Work is also underway on draft laws governing the legal ownership of weapons, which will bring much-needed clarity and control to the millions of trophy weapons in the country.1 With a new demobilization law yet to be submitted to parliament (at the time of writing) and relatively few veterans in Ukraine today, 2024 is a critical window of opportunity. Ukraine should use this time to determine and implement a comprehensive veterans policy before the wave of demobilizing veterans arrives. Although there was no upsurge in veteran-related organized crime in the 2014–2022 period, the size of the veteran population will be much larger. Even a fraction of these veterans falling into crime will have a significant impact on Ukrainian society. At present, there is little sign of institutional readiness. Implementation of the initiatives announced to date has been patchy and slow, and the Ministry for Veterans Affairs has lacked a permanent head for much of the first half of 2024. Demobilization, for those who are eligible, is a frustrating and even humiliating process. Our interviews with veterans revealed scarce access to information about benefits, with many unaware of what was available to them. In terms of rehabilitation, there is a lack of effective psychological, physiological, legal and social support for veterans. Some of this is explainable by the very real strictures the war has imposed on Ukraine: training that usually takes years must be completed in a matter of weeks.

But the need remains. Since the full-scale invasion, civil society organizations, many of which have worked in veterans affairs since 2014, have been making strenuous and effective efforts to cover the gaps in care and provision for veterans, from physiological and psychological support to forming business collectives and designing a ‘whole-life’ veterans policy. Yet many feel they are working in isolation, with the state resisting their attempts to work collaboratively and introduce strategic and innovative thinking to veterans affairs.2 Only the state can deliver a veterans policy with the scope and resources needed, but NGOs have much to offer, not least their agility and ability to deliver help where it is most needed. Both the state and civil society will be needed if Ukraine is to deliver a programme of ‘deep prevention’ – a strategy that addresses veterans’ needs at the level of contributing social factors rather than the individual level – which will reap the most dividends in reducing exposure to organized crime risks. The hard reality is that, as Ukraine dedicates as many resources as it can to the day-to-day conflict, it must also begin planning for the decades-long aftermath.

Geneva: Global Initiative Against Transnational Organized Crime , 2024 48p.

By Simon A. Cole Maurice Possley Ken Otterbourg Jessica Weinstock Paredes , Barbara O’Brien, Meghan Cousino, & Samuel R. Gross,

EXECUTIVE SUMMARY

A. THE CASES

• This report analyzes 151 cases in which defendants were exonerated between 1989 and 2023 in the United States and medicolegal death investigation (“death investigation” for short) contributed to the false conviction.

• The 151 exonerees lost a total of 1,837 years in prison, an average of 12.2 years per exoneree. That is less than the average of 14.6 years for exonerees convicted of comparable crimes but for whom death investigation did not contribute to the false conviction.

B. CASE CHARACTERISTICS

• Not surprisingly, 140 (93%) of the 151 cases in which death investigation contributed to the false conviction were homicides. However, death investigators did contribute to eleven non-homicide cases, all involving abuse of vulnerable people: children or dependent adults. Eight of these eleven were cases involving the Shaken Baby Syndrome (SBS) diagnosis, in which the top charge was child abuse.

• In more than one-third of the cases, the death investigation evidence consisted of a claim that the medical evidence was consistent with the prosecution’s theory of the crime, e.g., that the victim’s wounds were consistent with a weapon linked to the defendant.

• In another third of the cases the death investigation evidence concerned the cause of death.

• Manner of death and time of death evidence contributed to fewer cases.

C. DEMOGRAPHICS

• Women were overrepresented among the defendants for whom the death investigation contributed to their false conviction. Thirty-nine (26%) of the defendants in the 151 cases were female, more than three times the 8% of all exonerees who were female. Only around 5% of exonerees convicted of comparable crimes were female.-

• Relatedly, cases involving child victims were particularly vulnerable to contributions by death investigation. Nearly half (47%) of the 151 cases involved child victims. That compares to only 19% of all non-death-investigation exonerations and 34% of non-death-investigation exonerations for comparable crimes.

• Although concerns have been raised about racial bias in death investigation, the exonerees in death investigation exoneration cases were whiter than exonerees in general. One third of death investigation exonerees were Black compared to 53% of all exonerees. Similarly, 8% of death investigation exonerees were Hispanic, compared to 12% of all exonerees. The higher representation of whites diminishes somewhat if women are removed from the analysis.

D. DEATH INVESTIGATION SYSTEMS

• The United States has a patchwork death investigation system with variations among and within states. The two primary types are medical examiner and coroner systems. Most, but not all, experts perceive medical examiner systems to be superior and call for them to replace coroner systems. We did not find that more false convictions occurred under coroner systems. Instead, false convictions generally occurred in proportion to where more people live: their occurrence correlated with those counties’ and states’ proportions of the US population.

• Nor did we find that more false convictions occurred in systems with elected (rather than appointed) coroners and death investigators.

• In 22% of cases, the death investigation office that contributed to the false conviction was accredited by the National Association of Medical Examiners (NAME). Only 17% of US death investigation facilities are accredited.

E. QUALIFICATIONS OF DEATH INVESTIGATORS

• The highest qualification for death investigators in the US is generally considered to be board certification in the subspecialty of forensic pathology by the American Board of Pathology. However, for decades there have not been enough board-certified pathologists in the US to meet the need for death investigation services and autopsies. Therefore, many death investigations and autopsies are performed by less qualified personnel, such as pathologists without board certification, physicians with specialties other than pathology, and even, in some cases, non-physicians such as funeral directors. We did not find that most false convictions occurred in cases with underqualified death investigators. In fact, board-certified forensic pathologists contributed to 61% (92) of the 151 cases in this study.

National Registry of Exonerations (2024), 90p

By Christopher S. Carpenter and Hai V. Nguyen

Bans on menthol cigarettes have been adopted throughout the European Union, proposed by the US Food and Drug Administration, and enacted by legislatures in Massachusetts and California. Yet there is very limited evidence on their effects using real-world policy variation. We study the intended and unintended effects of menthol cigarette bans in Canada, where seven provinces banned them prior to a nationwide ban in 2018. Difference-in-differences models using national survey data return no evidence that provincial menthol cigarette bans affected overall smoking rates for youths or adults. Although menthol cigarette smoking fell for both youths and adults, youths increased nonmenthol cigarette smoking, and adults shifted cigarette purchases to unregulated First Nations reserves. Our results demonstrate the importance of accounting for substitution and evasion responses in the design of stricter tobacco regulations

Cambridge, MA:  NATIONAL BUREAU OF ECONOMIC RESEARCH, 2020. 66p.

By John P. Sullivan, Nathan P. Jones 

This chapter covers the early history of bandits including the role of bandits in revolution (e.g., Pancho Villa), the rise of urban guerrillas (e.g., Che Guevara and the Cuban Revolution, Carlos Marighella (the Minimanual of the Urban Guerrilla)) in Brazil and the Tupamaros in Uruguay, and the transition from ideological revolt to the criminal insurgency. The later transition will focus on the transition to crime by the FARC and ELN in Colombia succeeded by Bacrim in Colombia, mega-gangs in Venezuela, and Brazilian gangs (i.e., The Commando Vermelho and Primeiro Comando da Capital), followed by the rise of a criminal insurgency waged by territorial (third-generation) gangs (including maras such as MS-13 and Barrio 18 in Central America) and criminal cartels. The nexus between prison gangs and criminal gangs and the emergence of criminal enclaves (such as the Triple Frontier and Ciudad del Este) will be explored. A common thread will be a discussion of crime and social banditry (Hobsbawm) as mechanisms of revolt. Urban crime and instability and the emergence of crime wars and criminal insurgency will be discussed in context of state transition, globalization, and the rise of transnational organized crime.

London; New York: Routledge, 2021, 28p.

By Angélica Durán-Martínez

Objective/context: After the paramilitary demobilization in Colombia, Cali and Medellín—major cities with a long history of political and criminal violence—saw a decline in violence despite the continuing operation of criminal groups. Yet, while Cali remained more violent than the national average, with criminal groups displaying less ability to regulate local affairs and security, Medellín became less violent than the national average and criminal groups engaged in more intensive governance. Methodology: I compare these cities using case studies constructed through fieldwork and interviews with security officials, NGO and social leaders, and former members of groups, drawing on my long-term, ongoing research engagement in both locations. Conclusions: The contrast between these cities exemplifies varied manifestations of post-conflict criminality. I argue that this variation in criminal governance is connected to the wartime balance of power, specifically, the level of territorial control and political connections armed groups had before the demobilization. Where control and political connections of paramilitary before demobilization were high, post-conflict crime groups were more likely to engage in governance behaviors. By contrast, disputed territories during wartime were likely to experience less criminal governance. Drawing on path dependence ideas, I introduce an overlooked mechanism through which wartime orders affect criminal behavior: learning processes inside and outside criminal groups. Originality: I show that to fully understand post-conflict criminality, it is crucial to consider armed actors peripherally involved in the war but essential for territorial control, whereas to understand criminal governance, the mid-ranks of armed groups and learning processes are crucial.

Revista Colombia Internacional ,2024. 35p.

By Matthew J. Lindquist,  Eleonora Patacchin, Michael Vlassopoulos., Yves Zenou 

We study spillover effects within co-offending networks by leveraging the deaths of co-offenders for causal identification. Our results demonstrate that the death of a co-offender significantly reduces the criminal activities of other network members. We observe a decaying pattern in the magnitude of these spillover effects: individuals directly linked to a deceased offender experience the most significant impact, followed by those two steps away, and then those three steps away. Moreover, we find that the death of a more central co-offender leads to a larger reduction in aggregate crime. We also provide evidence consistent with a new theoretical prediction suggesting that the loss of a co-offender shrinks the future information set of offenders, altering their perceptions of the probability of being convicted and consequently affecting their criminal behavior. Our findings highlight the importance of understanding spillover for policymakers seeking to develop more effective strategies for crime prevention.  

  Bonn:: Institute of Labor Economics - IZA, 2024.

By Pamela B. Teaster; Erica Wood; Sally B. Hurme; E. Carlisle Shealy

This is the full report resulting from a four-part environmental scan of the issues of abuse by guardians and systemic guardianship abuse, requested in 2021 by the National Institute of Justice (NIJ), showing that more work and reform related to abuse and fraud by individual guardians and the guardianship system are required. The scan consisted of a literature review of research; a scan of the data landscape; a scan of legal, policy, and practice context; and considerations for data collection. This paper is the full report of the scan and its findings. On June 30, 2021, the U.S. Attorney General and the Secretary of the Department of Health and Human Services received a request for information from Senators Elizabeth Warren and Robert Casey about the roles of the Departments of Justice and Health and Human Services concerning the collection of data on adult guardianship — particularly data on abuse and fraud by guardians. Part 1 was an extensive literature review. Part 2 examined how states collect data on adult guardianship. Part 3 highlighted the difficulties in determining the number of adults with guardians as well as the prevalence of abuse by guardians. Building on this research, the authors recommended federal actions to help states detect and address abuse.

Unpublished report, 2022. 146p.

By John K. Roman, Anthony Washburn; Sofia Rodriguez; Caterina G. Roman; Elena I. Navarro; Jesse T. Brey; Benjamin M. Reist

This final report of the HAVEN Project (Harms After Victimization: Experience and Needs), which was launched in 2020 as an update to the 1996 National Institute of Justice study, Victim Costs and Consequences: A New Look. This document reports on the HAVEN Project’s use of data, measurement, and analytic tools that were not originally available in 1996; it also examines the HAVEN Project’s expansion of the taxonomy of harms from victimization and development of a survey instrument and methodology to facilitate the collection of self-reported data on the harms from violent victimization across multiple dimensions that are typically excluded from violent crime harms measurement. The HAVEN Project also introduced a regression-based cost-benefit model that may be integrated into causal models. The key research questions discussed in this final report are: if regression models of victimization harms, including estimates of the variance in harms experienced by victims, are feasible; how integrated data systems (IDS) can be employed to estimate trajectories of harms using the harm taxonomy developed, and what the strengths and limitations of those data systems; if those new definitions of direct and indirect harms to victims change estimated costs of victimizations, and if those new definitions are applicable to all crimes; if household survey data can generate estimates of the incidence of each victimization trajectory, and what can be learned from the analyses about those harms that may not require hospitalization; if new cost benefit analysis (CBA) methodologies can improve the quality of program evaluations that include harms to victims; and which types of victimization are most harmful.

Chicago: NORC at the University of Chicago, 2023. 37p.

By Janique Kroese, Wim Bernasco, Aart C. Liefbroer, Jan Rouwendal

Studies investigating the role of single-parent families in adolescent delinquency have seldom differentiated between types of single-parent families. Furthermore, they have typically assumed that parental disruption is a discrete event marking an abrupt change between dual-parenthood and single-parenthood. Using Dutch longitudinal population register data, we estimated fixed-effects panel models to assess (1) whether the event of parental disruption, either by parental separation or by parental death, increases subsequent adolescent delinquency and (2) whether parental disruption, either by parental separation or by parental death, has anticipatory, immediate, or delayed effects on adolescent delinquency. Our results showed that both parental separation and parental death seem to boost adolescent delinquency, and we found no difference between these types of single-parent families. However, when distinguishing between anticipatory, short-term, and long-term effects, we found a short-term increase in adolescent delinquency after a parental separation and an anticipatory reduction in adolescent delinquency before a parental death. Future research should pay more attention to diversity in the composition of single parent families, as well as to the anticipatory, short-term, and long-term consequences.

Journal of Developmental and Life-Course Criminology (2024) 10:288–308

By Benjamin Artz and  David M. Welsch:

We employ a unique 10-year panel dataset from California to examine both the effect crime has on homelessness as well as the effect homelessness has on crime. Our main estimator accounts for endogeneity by incorporating dynamics, controlling for time-invariant unobserved heterogeneity, and relaxing the strict homogeneity assumption for our key variables of interest. We find strong evidence that regions experiencing increases in property crime, but not violent crime, should expect a practically significant, 2024. 69p..in homelessness, whereas increases in homelessness increase the number of violent crimes, but not property crimes. Robustness and falsification checks confirm the results.

Bonn:  Institute of Labor Economics - IZA, 2024. 49p.

By Julia Tolmie, Rachel Smith, and Denise Wilson

How intimate partner violence (IPV) is conceptualized affects what we see when we look at situations involving IPV and what we think the solutions to the problem of IPV are—either in individual cases or in the development of broader legal and policy responses. In this article, it is suggested that while conceptualizing IPV as coercive control is an improvement over previous understandings, it does not go far enough. Coercive control must be located within a broader conceptualization of IPV as a form of social and systemic entrapment if it is not to operate in a harmful manner for victim-survivors.

Violence Against WomenVolume 30, Issue 1, January 2024, Pages 54-74

By Miguel Basto-Pereira, David P. Farrington, Laura Maciel

This work aims to investigate the role of sequences of risk factors from childhood to young adulthood in predicting subsequent criminal convictions. This study uses the Cambridge Study in Delinquent Development (CSDD) dataset, a prospective longitudinal research study that followed 411 males from South London from the age of 8 to 61 years. Temporal sequences of risk factors at ages 8–10, 12–14, and 16–18 were analyzed as predictors of subsequent criminal convictions up to the age of 61. Risk factors related to poverty, parenting problems, and children’s risk-taking predisposition at ages 8–10 emerged as prevalent starting points for the most highly predictive developmental sequences leading to convictions. The risk of a criminal conviction significantly increased if these risk factors were followed by low IQ scores or association with delinquent friends at ages 12–14, and by school and professional problems or drug addiction during late adolescence (ages 16–18). At each developmental stage, specific risk factors intricately combine to form chains of risk during development, subsequently predicting criminal convictions. A trajectory-of-risk-need-responsivity approach that identifies and breaks chains of risk factors that generate and enhance favorable conditions for criminal convictions is discussed.

Journal of Developmental and Life-Course Criminology (2024) 10:242–264

By Ido Kilovaty

The Colonial Pipeline ransomware attack, which shut down gas supply to the entire East Coast back in May 2021, has sparked debate as to the regulation of the pipeline’s cybersecurity. After ten years of inaction on the matter, the Transportation Security Administration (TSA) has issued two mandatory directives on pipeline cybersecurity. This Article delves into the propriety of the TSA as a pipeline security regulator, as well as the incomplete and ineffective approach currently laid out in the TSA’s pipeline cybersecurity directives. This Article argues that there may be other agencies more suitable for the task, such as the Federal Energy Regulatory Commission, acting under the auspices of the Department of Energy. It also provides specific recommendations as to the substance of any prospective pipeline cybersecurity regulation, such as the creation of more open-ended and flexible cybersecurity objectives as opposed to the current approach of prescriptive standards.

Houston Law Review, Vol. 60, 2023, Kilovaty, Ido, Cybersecuring the Pipeline (March 29, 2022). Houston Law Review, Vol. 60, 2023,

By Simon D. Venema, Marieke Haan, Eric Blaauw, René Veenstra

Little is known about the conditions under which paternal incarceration is harmful to children and the mechanisms that explain this. This study addressed the family relationship context in the associations between paternal incarceration and adolescents' internalizing and externalizing problem behaviors. Using data from the Future of Families and Child Wellbeing Study, a moderated mediation model was specified where paternal incarceration predicted adolescents’ internalizing and externalizing problem behaviors through family relationship quality, and where the mediating role of family relationship quality was moderated by pre-incarceration family relationship characteristics. Using latent profile analyses, three pre-incarceration family clusters were identified (“Cohesive”; “Fragmented”; “Disharmonious”). Analyses indicated that the association between paternal incarceration and family relationship quality differed across pre-incarceration family clusters and that decreased father mother relationship quality mediated the negative association between paternal incarceration and adolescents’ internalizing and externalizing problem behaviors among “Cohesive” and “Fragmented”, but not among “Disharmonious” family clusters. The findings suggest that adolescents with more harmonious pre-incarceration family relationships are most vulnerable to the negative consequences of paternal incarceration. The study demonstrates the need to consider the family relationship context to understand the intergenerational consequences of incarceration.

Journal of Developmental and Life-Course Criminology (2024) 10:213–241

By Corrie Williams, Tara Renae McGee, Shannon Walding, Christine E. W. Bond

While conduct problem behaviour initiated in early childhood often escalates in frequency and seriousness through adolescence, a notable deceleration is typically seen by mid-adolescence. It has been hypothesised that prosocial behaviour, characterised by acts like sharing and comforting, may play a role in this deceleration. However, there is a distinct gap in the current literature when it comes to understanding the temporal dynamics between the acceleration of prosocial behaviours and the deceleration of conduct problem behaviour. This study seeks to bridge this gap. Using a General Cross-Lagged Panel Model (GCLM) and data from the Longitudinal Study of Australian Children (LSAC), we investigated temporal dynamics and sequence of how the acceleration of prosocial behaviour influences the deceleration of conduct problem behaviour between ages 4 and 15. Results indicate that increases in prosocial behaviour facilitate the deceleration of conduct problem behaviour, with increases in prosocial behaviour preceding decreases in conduct problem behaviour. Further, we show a cumulative effect of increases in prosocial behaviour on decreases in conduct problem behaviour over time. This knowledge provides a foundation for understanding how timely prevention and intervention strategies that include the mechanisms for increasing prosocial behaviour may interrupt the con duct problem behaviour trajectories of children and adolescents.

Journal of Developmental and Life-Course Criminology (2024) 10:169–192

By Ido Kilovaty

Generative AI platforms, like ChatGPT, hold great promise in enhancing human creativity, productivity, and efficiency. However, generative AI platforms are prone to manipulation. Specifically, they are susceptible to a new type of attack called “prompt injection.” In prompt injection, attackers carefully craft their input prompt to manipulate AI into generating harmful, dangerous, or illegal content as output. Examples of such outputs include instructions on how to build an improvised bomb, how to make meth, how to hotwire a car, and more. Researchers have also been able to make ChatGPT generate malicious code. This article asks a basic question: do prompt injection attacks violate computer crime law, mainly the Computer Fraud and Abuse Act? This article argues that they do. Prompt injection attacks lead AI to disregard its own hard-coded content generation restrictions, which allows the attacker to access portions of the AI that are beyond what the system’s developers authorized. Therefore, this constitutes the criminal offense of accessing a computer in excess of authorization. Although prompt injection attacks could run afoul of the Computer Fraud and Abuse Act, this article offers ways to distinguish serious acts of AI manipulation from less serious ones, so that prosecution would only focus on a limited set of harmful and dangerous prompt injections.

Loyola of Los Angeles Law Review, Vol. 58, 2025, Kilovaty, Ido, Hacking Generative AI (March 1, 2024). Loyola of Los Angeles Law Review, Vol. 58, 2025,