FEBRUARY - SEPTEMBER 2025
Brandon May
Background: As generative artificial intelligence (GenAI) advances into law enforcement practice, its use in investigative interviewing raises conceptual, evidentiary, and ethical challenges. This commentary posits the phenomenon of DrAIfting – the epistemic and procedural drift that occurs when investigators defer to AI-generated prompts, scaffolds, or hypotheses in ways that erode professional judgment, rapport, and trauma-informed practice. Drawing on empirical evidence from cognitive psychology, forensic interviewing, and recent studies of human–AI interaction, I highlight risks of memory contamination, metacognitive deskilling, and the privileging of procedural exhaustiveness over adaptive relational judgment. I also examine the legal implications of deploying large language models (LLMs) in policing, noting that such systems currently fail to meet Daubert criteria of testability, transparency, and general acceptance. In addition, I discuss that current AI-systems (e.g., avatar-based simulations) and constrained administrative applications may offer pedagogical or clerical value; however, their integration must remain bounded by ethical oversight, empirical validation, and transparent governance. Without such safeguards, GenAI risks reshaping investigative interviewing into a mechanistic process that undermines evidentiary reliability, voluntariness, and moral integrity. Instead, I argue that AI should be positioned as a supplementary tool, reinforcing rather than displacing the human virtues, judgment, and dialogical skills at the core of professional interviewing.
Key Words: Artificial Intelligence; Large Language Models; Investigative Interviewing; Memory; Rapport
Background: Interviews with suspects, victims, and witnesses are among the most important and regular tasks undertaken by police/law enforcement agencies to progress criminal investigations. The present article addresses a critical gap both in the literature and practice of investigative interviewing—namely, the lack of a coordinated global action to establish and implement a universal standard. While countries like England and Wales, Norway, and Australia have successfully adopted rapport-based practices, these efforts remain largely confined to individual national contexts. Broader collective efforts involving academics, activists, and practitioners worldwide—focused on expanding the adoption of investigative interviewing—are scarce. To address this gap, the article outlines developments in the last few years that have attempted to provide a more universal approach to investigative interviewing and introduces a pioneering global action to support its implementation. The initiative and action are justified by their potential to improve global consistency, fairness, and effectiveness in investigative practices while fostering international collaboration. Its significance lies in broadening the reach of investigative interviewing, improving justice outcomes globally, and establishing a framework for sustained cooperation and knowledge exchange.
Key Words: Mendez Principles, ImpleMéndez, Investigative Interviewing; Coercive interrogation; Criminal investigation
William O’Donohue, Sneha Gupta, Brandon T. Hunley
Background: Forensic interviews (FI) with adults who allege sexual assault provide essential information for subsequent legal decisions that impact the interviewee and the alleged perpetrator. FIs also play an essential role in identifying offenders and providing justice for survivors. However, FIs can vary in quality. While criticisms regarding FIs exist, the literature lacks a comprehensive model of interview errors. We present 32 FI errors and the rationale for characterizing each of these as errors.
This preliminary typology can assist researchers in assessing the impact of errors, in finding ways to avoid these errors, as well as making improvements in this enumeration of interview errors.
.
Key Words: Forensic interviews, sexual assault, rape, suggestibility, rape myths, errors, content validity
Background: For the information elicited within an investigative interview to be admissible in court in many Western countries, the methods used to obtain it must not violate fundamental principles of justice or fairness (i.e., “shock the conscience of the community”). Despite the frequent application of this principle in judicial systems globally, little research exists in which public perceptions of interviewing tactics have been examined. The current research therefore sought to fill this gap in the literature by examining how the public perceives various tactic types (rapport-based, overt, psychological) and if this may differ depending on the type of interview (i.e., investigative vs. intelligence-gathering). Across two studies, participants (N = 122, N = 133) read transcripts in which a crime had occurred in the past or was believed to occur in the future (the same day or in a week), then provided ratings regarding the effectiveness and appropriateness of various tactics characteristic of three different approaches (overt, psychological or rapport-based). Across perceptions of effectiveness and appropriateness, findings indicated a clear preference for rapport-based tactics. Implications for the “shock the conscience” principle and suspect interviewing more broadly are discussed.
Key Words: Investigative interview, shock the conscience, intelligence gathering, rapport, public perceptions
Joseph Eastwood
Background: This op-ed revisits the longstanding debate surrounding the Reid Technique in suspect interviewing, nearly fifteen years after the author’s original exchange on the topic. Prompted by a recent Canadian case where a coerced confession led to a collapsed homicide trial, the piece examines whether the Reid Technique is truly the source of problematic interviewing practices, or whether its name is too often used as a convenient scapegoat. The author argues that while Reid contains deeply flawed elements—such as reliance on behavioral cues, presumption of guilt, and restrictions on dialogue—it is not responsible for all instances of poor practice. Instead, careless or unethical interviewing can arise independently of any prescribed method. The article calls for greater precision in criticism, urging academics and practitioners alike to engage with the Reid Manual directly rather than perpetuating mischaracterizations. Moving forward, the author highlights evidence-based, rapport-driven approaches as more promising alternatives, and emphasizes the importance of continued collaboration between researchers and practitioners to embed ethical, effective, and scientifically grounded interviewing practices.
Key Words: Reid Technique; Investigative Interviewing; Coercive Interrogation; Suspect Confessions; Evidence-based Practices; Rapport-based Approaches
Background: This exploratory research aimed to examine nonverbal behaviours exhibited by interviewers within a virtual interviewing context and (1) their relation to interviewees’ perception of rapport and subsequent information disclosure, and (2) external observers’ perception of rapport. Participants (N = 66) were shown a sexual education video and then questioned about the video in a live virtual interview using either a Rapport (e.g., smiling, nodding, looking towards the screen) or Limited Rapport (e.g., no smiling, looking away) approach. In Study 1, we coded the interviews for five interviewer nonverbal behaviours (i.e., smiling, nodding, looking away, shaking the head, and touching hair or face) and examined whether they were associated with increased information disclosure from interviewees and their perception of rapport. In Study 2, we showed the substantive phase of 19 of these interviews to six research assistants, blind to the study, to examine if external observers’ ratings of rapport match that of the interviewee and what information these external observers report using in generating their rapport ratings. Study 1 suggests that interviewees detected interviewers’ nonverbal behaviours, with nodding linked to higher perceptions of rapport and both smiling and nodding linked to greater information disclosure, though nodding was also related to increased incorrect details. Study 2 suggests that external observers used nonverbal cues to rate rapport, but their ratings only moderately matched interviewees’, underscoring the role of context and perspective in rapport assessment. Implications regarding the role of nonverbal behaviours in virtual investigative interviews are discussed.
Key Words: Virtual Interviews; Investigative Interviews; Rapport; External Observers; Information Disclosure; Nonverbal Behaviours
Sarah L. Deck, Sonja P. Brubacher, Martine B. Powell
Background: There is evidence that an elevated proportion of adults convicted of sexual offences against children are autistic. When investigators seek to gather information from these offenders, the risk for miscommunication is high. There is limited guidance on the most effective questioning styles for investigators to use with autistic offenders. In the current study, the utility of best-practice interviewing techniques for this cohort was explored. Nine investigators who regularly monitor persons on the Child Protection Offender Registry were trained in autism awareness and general vulnerable witness interviewing best-practices. Subsequently, they were interviewed about their perceptions of the interview training with regards to its applicability for questioning autistic offenders. The findings provide insight into strategies that may be particularly helpful for establishing rapport with autistic respondents, namely the use of structured open-ended questions and narrative practice. The findings also complement previous quantitative research on questioning techniques that are effective with autistic respondents, suggesting that appropriate scaffolding is necessary. However, the potential disadvantages of asking specific questions with reluctant autistic respondents was also highlighted. These results suggest promising avenues for enhancing the quality of information-gathering conversations with autistic respondents, which warrant further investigation through systematic quantitative research.
Key Words: Autism Spectrum Disorder, Investigative interviewing, Open-ended questions, Rapport, Narrative practice
Background: In child abuse cases, the testimony of child victim witnesses often stands as the sole evidence. Some responsibility for ensuring the quality of this testimony falls on those who conduct preliminary conversations prior to criminal proceedings. Such conversations must be conducted professionally.
Objective: As one of the largest group of professionals reporting child abuse, teachers often start the initial conversations with children in cases of child sexual abuse (CSA).
Participants and setting: This interview study examined teachers’ attitudes and knowledge about communicating with children in cases of suspected CSA, and sought advice from child protection and law enforcement professionals for teachers and school professionals.
Methods: Semi-structured expert interviews were conducted with 45 professionals from Germany. The interviews were analyzed using qualitative content analysis.
Results: The findings of this study align with previous international research, underscoring the need for more comprehensive teacher education. A need for more education on talking to children for school professionals became evident. Furthermore, school was viewed as an ideal place for intervention and support for children in CSA cases.
Conclusion: Developing a training and education program on how to handle cases of suspected CSA to support children in schools is deemed highly important.
Key Words: teacher training; child sexual abuse; interview study; multi-professional; school