Open research questions in Deception detection and forensic psychology

What the literature leaves open

• These results open interesting avenues for further research. First, future research should investigate the relationship between truth, oath and incentives to lie and later memory performance. Research is needed to determine whether the incentives to lie associated with an oath simply affect what the participant recalls (biased responding), or whether it affects their memory or source attribution of the original event. If participants’ errors in the first phase of the test are due to social influence (incentive to lie), then errors in a later memory task, when incentives to lie are removed, could be due to false memories. Source monitoring theory (Johnson et al., 1993) suggests that individuals engage in a meta-cognitive process of identifying the source of their memories, and that source attribution errors (i.e., when an individual believes that a memory comes from a lived experience when it actually comes from suggestion or imagination) may be one of the causes of false memories. An oath to tell the truth in the truth condition would encourage participants to evaluate the information in their memory. After a delay, they might be better able to distinguish the source of their memories. Moreover, taking an oath may strengthen the confidenceaccuracy relationship, which in turn could enhance a witness’s perceived credibility in court (Tenney et al., 2007). In contrast, when participants are incentivized to lie, choosing the wrong answer in the first phase could increase access to the false memory trace at the expense of the original information, and even more so under oath (for a systematic review of the effect of lying on memory contamination, see Battista & Otgaar, 2022). Further research should also explore the impact of taking an oath in evidence– based interviewing. During investigative interviews, particularly in the free account phase, individuals engage in a strategic regulation process, assessing the accuracy and relevance of the information beforehand deciding whether to verbalize it or not (Koriat & Goldsmith, 1996). An oath may encourage more careful strategic regulation, potentially enhancing memory performance in truthful accounts. However, it is also likely to increase the decision criterion of participants, making them more cautious in their responses and leading to a tainted-truth similar effect (for a discussion, see Blank & Launay, 2014). Since verbal lie detection tools rely on prompted recall techniques (e.g., encouraging interviewees to say more, the Verifiability Approach, Strategic Use of Evidence), it would be relevant to test how an oath would alter the recall of deceptive individuals (see Vrij et al., 2022, for a review).

  Does the oath enhance truth-telling in eyewitness testimony? experimental evidence · 2026 · DOI

• Our findings are mixed, but we find that images of White male defendants were linked with larger amplitudes on one ERP that has been associated with expectation violation and one ERP that has been associated with more “effortful” and “deep” processing of facial stimuli.

  Defendant race and neuropsychological response: an exploratory study · 2026 · DOI

• This study is not without limitations. First, the multimodal feature analysis relies on 16 basic multimodal features, which may not fully capture the audiovisual elements in short videos. Future research may expand the scope of this study and incorporate more platform-specific features and filters to provide a more comprehensive understanding of multimodal features of unsafe content. Second, the generalizability of the study’s findings is lim- ited by the dataset in two ways: (a) some short videos be- came unavailable during the study due to takedowns resul- tant from internal content moderation or changes in privacy settings, and (b) thematic analysis was conducted only on English-language captions. On the first point, only 4.1% (n = 186/4,496 URLs) of videos were inaccessible. Because the purpose of the study was to identify multimodal features of unsafe videos that may lead to subsequent negative impacts over time, we think this limitation is a minor point. On the second point, only 5.5% (n = 93/1,696 captions) of the cap- tions are in non-English languages. Future work needs to take into consideration all languages in texts, and culture- or language-specific multimodal features to expand the global relevance and generalizability of this work (Lai, Qie, and Rau 2021). Third, although this study analyzes short video content within the context of an algorithm auditing experiment, this study is observational in nature. As such, it provides correla- tional evidence linking multimodal features of short videos to unsafe content, without addressing the question of how such features influence subsequent information processing and psychological or behavioral changes. Future research can build on these findings using longitudinal designs or ex- perimental approaches to better understand how unsafe mul- timodal contents evolve and affect children and teenagers. Lastly, engagement with short videos was measured as the number of views, likes, and comments, but the data on dis- likes or downvotes was unavailable. This is because only YouTube Shorts allows dislikes but the platform does not report them separately. The like count reflects the net value after subtracting dislikes. This lack of negative engagement metrics limits our ability to capture more nuanced audi- ence responses, especially for unsafe content that is widely viewed but potentially negatively received. Future research could incorporate alternative metrics, such as watch dura- tions and shares, to better evaluate audience engagement at the individual level.

  Catching Dark Signals in Algorithms: Unveiling Audiovisual and Thematic Markers of Unsafe Content Recommended for Children and Teenagers · 2026 · DOI

• INVESTIGATIVE ACTIONS USING ARTIFICIAL TOOLS INCREASING RELIABILITY IN VERBAL INTELLIGENCE Nurmukhammed Tazhigulov (a) Serguei Cheloukhine (e) Yevgeniy Shulgin (b)1 Larissa Kussainova (c) Ainura Omarova (d) (a)PhD Student, Academy of Law Enforcement Agencies at the Prosecutor General’s Office of the Republic of Kazakhstan; E-mail: [email protected] (b)Professor, School of Law, Karaganda Academy of the Ministry of Internal Affairs of the Republic of Kazakhstan named after B.S. Beisenov, Karaganda, Kazakhstan; E-mail: [email protected] (c)Professor, School of Law, Karaganda National Research University named after academician Ye.A.Buketov, Karaganda, Kazakhstan; E-mail: [email protected] (d)Professor, School of Management, Karaganda National Research University named after academician Ye.A.Buketov, Karaganda, Kazakhstan; E-mail: [email protected] (e)Professor, John Jay College of Criminal Justice, New York, USA; E-mail: [email protected] A R T I C L E I N F O A B S T R A C T Article History: Received: 12th October 2025 Reviewed & Revised: 12th October 2025 to 11th April 2026 Accepted: 12th April 2026 Published: 15th April 2026 Keywords: Artificial Intelligence, Blockchain, Evidence Reliability, Verbal Investigative Actions, Digitalization of Justice JEL Classification Codes: K10, K24, K40, K14 Peer-Review Model: External peer review was done through double-blind method. Kazakhstan has prioritised digitalising its criminal justice processes, implementing AI-driven tools such as automated forensic systems and "digital autopsy" platforms to increase efficiency and transparency. However, challenges persist: interrogation recordings can still be falsified or incomplete due to inadequate technical equipment and procedural gaps, undermining evidence integrity. This study investigates a combined AI-blockchain framework to enhance the reliability of verbal investigative actions. It examines an integrated system in which automatic speech recognition (ASR) transcribes interrogation audio, and a private blockchain ledger immutably timestamps each transcript. The study employs a mixed-methods approach: data include recorded criminal interrogation sessions (audio transcripts) and relevant legal documents. A neural ASR model was trained on these transcripts (approximately 50 sessions) to produce digital text, and standard analytical methods were applied to evaluate system performance. Simultaneously, each transcript was recorded on a blockchain registry, and metrics such as transcription accuracy and error rate were measured. Results show the proposed system attains high transcription accuracy (94% word recognition) and significantly reduces manual review time (by about 40%).

  PROSPECTS FOR INCREASING RELIABILITY IN VERBAL INVESTIGATIVE ACTIONS USING ARTIFICIAL INTELLIGENCE TOOLS · 2026 · DOI

• In general, police under-weighed evidence in the RMP format and were more accurate at judging suspect guilt when the evidence was in the LR format and when the evidence was inconclusive.

  Police evaluation of evidence: statistical format and evidence type · 2024 · DOI

• Arkes and Koehler's argument implicitly relies on the assumption that inconclusive decisions are truly a response by examiners to samples lacking information sufficient to reach more definitive conclusions, and thus will distribute relatively evenly between same-source and different-source comparisons.

  Signal detection theory fails to account for real-world consequences of inconclusive decisions · 2022 · DOI

• Based upon a critical analysis of the MS literature, we argue that this tool is not currently ready for practical usage, as its utility has not been fully established.

  A critical analysis of the Model Statement literature: Should this tool be used in practice? · 2020 · DOI

• Aims of the paper (scientific and/or social): This empirical research aims to further examine the possible existence of gender-related effects in eyewitness identification performance in the rarely studied culture of Bosnia and Herzegovina, and thus to determine implications for eyewitness identification in law enforcement practice there and elsewhere.

  Gender Effects Regarding Eyewitness Identification Performance · 2020 · DOI

• Summary Wh‐ prompts (what, how, why, who, when, and where) vary widely in their specificity and accuracy, but differences among them have largely been ignored in research examining the productivity of different question types in child testimony.

  The Productivity of Wh‐ Prompts when Children Testify · 2016 · DOI

• The author of the present paper discusses the most important points of this debate, showing how the low level of theoretical maturity, the lack of consensus on the rules of applying research techniques and interpreting results, and the unrealistic demands of editors of empirical journals may have contributed to this crisis.

  Post-Stapelian psychology. Discussions on the reliability of data and publications in psychology · 2015 · DOI

• Insight into these characteristics will help law enforcement personnel in two ways: It provides insight into who would be suitable for undercover operations, and it may help lie detectors, because one reason why people make errors in lie detection is that they do not take the full complexity of deception into account and seem to have limited knowledge about what is actually going on in a liar's mind.

  Good Liars · 2010 · DOI

• This study examined a sample of 142 of the more demanding interviews, (both from audiotapes and transcripts of actual interviews with suspects) and it revealed that, while investigators generally displayed ethical interviewing standards and tended to use open questioning techniques, few interviews were skilled.

  Interviewing Suspects of Fraud: An In-Depth Analysis of Interviewing Skills · 2010 · DOI

• A meta-analytic review of research comparing biased and unbiased instructions in eyewitness identification experiments showed an asymmetry; specifically, that biased instructions led to a large and consistent decrease in accuracy in target-absent lineups, but produced inconsistent results for target-present lineups, with an average effect size near zero (Steblay, 1997).

  A re-examination of the effects of biased lineup instructions in eyewitness identification. · 2005 · DOI

• Whereas substantial research has investigated adolescent information management and lying, little is known about the development of adolescents' reasoning about parents' covert monitoring and lying.

  Early Adolescents' Reasoning About Secretive Parental Monitoring and Lying: Assessing Covert Strategies for Monitoring Risky Behaviors · 2026 · DOI

• While research on romance scams is expanding, victims lived experiences of recognizing and responding to deception in mobile-mediated interactions remain insufficiently understood.

  When Suspicion Becomes Detection. Folk Deception Cues and Detection Strategies in Online Dating Romance Scams · 2026

• A limitation of this study is the reliance on global discrepancy scores rather than frame- by-frame time-series analysis. As we were limited to aggregated metric data in the current study, we could not perform retrospec- tive frame-by-frame synchronisation.

  Unmasking Deception: Examining Receivers’ Emotional Responses with Automated Facial Expression Analysis in Veracity and Valence · 2026 · DOI

• Despite the growing importance of audio in forensic settings, to date there are no standardized guidelines for transcribing intercepted speech materials or police interviews.

  Forensic Transcription and the Challenge of Segmental and Suprasegmental Cues · 2025 · DOI

• But except changing topics, the other three types of management formats can all provide suspects with opportunities to alter their original claims of insufficient knowledge.

  Police officers’ management of suspects’ <i>I don’t know</i> responses in Chinese investigative interviews · 2024 · DOI

• However, little is known about which behavioral cues elicit perceptions of authenticity in others and whether interviewees who exhibit such authenticity cues are more likely to perform better in the job interview and on their actual job.

  Observing Interviewees’ Inner Self: How Authenticity Cues in Job Interviews Relate to Interview and Job Performance · 2024 · DOI

• Research indicates that crime-first language ("criminal") increases stigma, but there is limited evidence comparing person-first language ("person with a conviction") to other non-deviant terminology.

  What’s in a Label? Public Use and Perceptions of Labeling Alternatives in Criminology · 2024 · DOI

• While live-link interviews have been successful in legal and medical contexts, its potential for eyewitness interviews remains insufficiently studied, particularly in terms of non-verbal misinformation.

  The Digital Witness: Exploring Gestural Misinformation in Tele-Forensic Interviews with 5-8-Year-Old Children · 2024 · DOI

• We conducted a systematic review showing that pre-event instructions, instructions given before event exposure, are rarely reported and those that are reported vary in the extent to which they warn participants about the nature of the event or tasks.

  The effect of pre-event instructions on eyewitness identification · 2023 · DOI

• Despite the potential significance of stress on individuals during investigative interviews, the examination of its general impact remains an underexplored area of research, with previous studies primarily focused on the specific phenomenon of stress-induced false confessions.

  Exploring Interrogation-Related Stressors: Factors Influencing Apparent Stress in Investigative Interviews with Suspects · 2023 · DOI

• Notably, both studies were limited by the use of an experimental design that was intended to capture initial impressions rather than fully mimic standard assessment and courtroom processes.

  Perceived Credibility of Social Media Data as a Collateral Source in Criminal Responsibility Evaluations Using an Experimental Design · 2021 · DOI

• The approach leads to a broader definition of lying that can account for lies that are told while speaking non-literally or with the help of presuppositions, and it allows for a parallel definition of misleading, which so far is lacking in the debate.

  The Lying-Misleading Distinction: A Commitment-Based Approach · 2021 · DOI

Most-cited papers in Deception detection and forensic psychology

• Tactical deception in primates · Behavioral and Brain Sciences · 1988 · 552 citations
• The evolution and psychology of self-deception · Behavioral and Brain Sciences · 2011 · 434 citations
• Variance in faking across noncognitive measures. · Journal of Applied Psychology · 2000 · 270 citations
• AN EXPLORATION OF DECEPTION AS A COMMUNICATION CONSTRUCT · Human Communication Research · 1974 · 177 citations
• A comfortable witness is a good witness: rapport‐building and susceptibility to misinformation in an investigative mock‐crime interview · Applied Cognitive Psychology · 2011 · 140 citations
• When emotions get the better of us: the effect of contextual top-down processing on matching fingerprints · Applied Cognitive Psychology · 2005 · 136 citations
• The effects of repetition frequency on the illusory truth effect · Cognitive Research Principles and Implications · 2021 · 136 citations
• When lovers become leery: The relationship between suspicion and accuracy in detecting deception · Communication Monographs · 1990 · 118 citations
• The role of psychophysiology in forensic assessments: Deception detection, ERPs, and virtual reality mock crime scenarios · Psychophysiology · 2007 · 103 citations
• Can we reduce facial biases? Persistent effects of facial trustworthiness on sentencing decisions · Journal of Experimental Social Psychology · 2020 · 80 citations

Most recent work

• Artificial Intelligence–based investigation of filler selection strategies. · Law and Human Behavior · 2026
• Mapping the Information Flow Within Deceptive Messages: A Test of the IM Propositions of IMT2 · Journal of Language and Social Psychology · 2026
• The Cases Against T - An Exploration of Risk Factors for False Confessions in a Danish Police Investigation and Subsequent Court Proceedings · Journal of Forensic Psychology Research and Practice · 2026
• A Scoping Review of the Literature on Embedded Truths and Lies in Investigative Research · Cognition Brain Behavior An Interdisciplinary Journal · 2026
• PROSPECTS FOR INCREASING RELIABILITY IN VERBAL INVESTIGATIVE ACTIONS USING ARTIFICIAL INTELLIGENCE TOOLS · Bangladesh Journal of Multidisciplinary Scientific Research · 2026
• Deception behind bars: A mixed‐methods investigation of inmates' beliefs about lie detection · Legal and Criminological Psychology · 2026
• Tracking Concealed-Information Signatures Over Time: Passive EEG Decoding of Probe vs. Irrelevant Trials · JoCN Forum · 2026
• Public perceptions of police use of artificial intelligence to assist in generating police reports: findings from an experimental study · Policing and Society · 2026
• Implementation and Validation of “SER-DesVi”: An Empirical System for Assessing the Police Risk of Injury, Suicide and Homicide in Cases of Missing Persons · European Journal on Criminal Policy and Research · 2026
• Catching Dark Signals in Algorithms: Unveiling Audiovisual and Thematic Markers of Unsafe Content Recommended for Children and Teenagers · Proceedings of the International AAAI Conference on Web and Social Media · 2026

Related topics in Psychology