Exploring automation bias in human–AI collaboration: a review and implications for explainable AI
Open accessThe DOI matches its Crossref record; 75 of the 75 references it cites were found in a registry or an index; 17 of the 264 sources its claims cite were checked for support.
Abstract
Systematic review of automation bias in high-stakes human-AI collaboration, examining cognitive mechanisms, expertise, trust dynamics, verification demands, and the limits of explainable AI for mitigating overreliance.
Overview
CoChat Summary
## TL;DR
This paper is a PRISMA-based review of automation bias in human-AI collaboration, synthesizing 35 peer-reviewed experimental studies from 2015–April 2025. It concludes that trust, task complexity, AI literacy, expertise, and individual differences shape overreliance, while explainability/transparency alone often do not prevent automation bias and can sometimes worsen it.
## Key findings
- The review included 35 studies, published across 21 journals, with a combined 19,774 participants; all included studies used quantitative, experimental methods.
- Trust and task complexity emerged as the main drivers of automation bias; the paper reports that trust accounted for up to 24.1% of the variance in reliant behavior in one study (Küper et al. 2025).
- Explainability was mixed or limited: CFFs reduced reliance on incorrect AI recommendations compared with simple XAI, but were seen as more cognitively demanding; multiple explanation types did not improve detection of incorrect AI recommendations, and trust calibration feedback did not help.
- The review reports several quantitative examples of bias effects, including Bond et al. (2018) where incorrect automated diagnoses reduced ECG interpretation accuracy from 84.86% to 41.66% for fellows and from 86.38% to 27.43% for non-fellows, and Dratsch et al. (2023) where mammography accuracy dropped from 79.7% to 19.8%, 81.3% to 24.8%, and 82.3% to 45.5% across experience groups when AI predictions were incorrect.
## Methods
- Systematic review conducted using PRISMA 2020; the authors defined five research questions on causal/mediator factors, co-occurring biases, explainability, AI literacy/mental models, and individual differences.
- Searches were run in SCOPUS, ScienceDirect, PubMed, and Google Scholar in May 2025, using peer-reviewed English-language studies from January 2015 to April 2025 that were quantitative, experimental, and focused on automation bias or over-reliance.
- Screening proceeded from title/abstract to full text, with 35 studies included; the two reviewers reported 0.95 Cohen’s kappa agreement.
## Limitations
- The paper states that restricting inclusion to peer-reviewed journal articles from 2015–2025, English-language studies, and experimental designs may have excluded earlier work, non-English studies, preprints, qualitative studies, and mixed-methods research.
- The paper states its search strategy lacked sensitivity and missed some relevant studies due to indexing gaps and terminology variation.
- The paper notes limited evidence on cultural and socio-demographic factors because of the current lack of cross-cultural comparison studies.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
17 Valid · 3 Not valid · 203 Not checked · 48 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“To the best of our knowledge, existing reviews have focused on AB in relation to traditional computer-based aids rather than AI-driven systems (see Goddard etal. 2012; Lyell and Coiera 2017).”
What the sources say
- Not matchedcites Goddard etal. 2012
An author–year citation the paper’s reference list doesn’t settle. We don’t search these by name.
- Partly supportedcites Lyell and Coiera 2017 matched to “Automation bias and verification complexity: a systematic review” (2017) sourceExperimental — accuracy not yet measured
The Introduction and Method describe a review of automation bias involving clinical decision support and other automated aids, which supports the characterization of this review's focus. Our copy does not distinguish those aids from AI-driven systems or establish that existing reviews generally excluded AI-driven systems. (LLM, conf=0.82)
hedged- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Rests on an assumption not shown here.
- Not matched
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Cognitive overload is considered the most significant factor inhibiting users from adequately verifying automation accuracy, thereby increasing AB:”
What the sources say
- Partly supportedcites Lyell and Coiera 2017 matched to “Automation bias and verification complexity: a systematic review” (2017) sourceExperimental — accuracy not yet measured
The Discussion hypothesizes that cognitive overload can prevent users from adequately verifying automation and lead to automation-bias errors. It does not establish cognitive overload as the most significant factor; the Conclusion says further research is needed to test its impact. The supplied passages do not include page numbers. (LLM, conf=0.88)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Users often fail to critically evaluate automation outputs when they perceive the system to be highly reliable (Lyell and Coiera 2017).”
What the sources say
- Supportedcites Lyell and Coiera 2017 matched to “Automation bias and verification complexity: a systematic review” (2017) source
The Introduction (page unavailable) says high system accuracy may foster trust and that users with greater trust in automation are less likely to detect its failures. It also describes automation bias as replacing vigilant information seeking with reliance on automated cues, supporting the paraphrase. (LLM, conf=0.90)
plainstates a possible cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“the harder the verification complexity, that is the complexity of verifying that the automation is performing correctly, the greater the likelihood of undue trust in the system (Lyell and Coiera 2017).”
What the sources say
- Partly supportedcites Lyell and Coiera 2017 matched to “Automation bias and verification complexity: a systematic review” (2017) sourceExperimental — accuracy not yet measured
The Discussion (page unavailable) says high verification complexity appears to increase the risk of automation bias, and the Method defines verification complexity as the difficulty of checking whether automation performs correctly. The claim presents that tentative relationship as a definite one; the paper does not establish that likelihood rises consistently as complexity increases. (LLM, conf=0.83)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“In line with these results, Dikmen and Burns (2022) found that providing domain-specific knowledge helped non-expert users to critically evaluate AI outputs, relying less on incorrect AI predictions.”
What the sources say
- Supportedcites Dikmen and Burns 2022 matched to “The effects of domain knowledge on trust in explainable AI and task performance: A case of peer-to-peer lending” (2022) source
Section 6.1 (page unavailable) reports that participants given domain-relevant guidance made fewer investments when the AI underpredicted loan-default risk. It says they may have noticed discrepancies and questioned the AI’s guidance, and concludes that conveying domain knowledge reduced over-reliance on incorrect AI predictions. The finding concerns the peer-to-peer lending task studied. (LLM, conf=0.91)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
132 more claims · 9 with their cited sources checked
Create a free account to see all 271 claims and every source verdict.
- It may state two separate claims as one. 96
- Probably not a substantial claim. 55
- Borderline: one of the tests for a substantial claim is unclear. 2
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“A comparable dynamic may plausibly extend to generative AI used as an information source, where fluent and confident outputs can be accepted uncritically (Romeo and Conti 2026), although this transfer across system classes should be treated with caution.”
From Epistemic Trust in Generative AI as an Information Source: Development and Validation of the Trust in AI-Epistemic Scale (TAI-E) (2026) · cites this paper as Romeo and Conti 2026
In the citing paper
“This configuration does not fit either of two simplified narratives that frequently appear in the surrounding literature: the assumption that analytic engagement necessarily reduces trust (Alon-Barkat and Busuioc 2023; Romeo and Conti 2026;Stanovich et al. 2016), or the assumption that trust primarily reflects uncritical acceptance (Davis 1989;Hoff and Bashir 2015).”
From Epistemic Trust in Generative AI as an Information Source: Development and Validation of the Trust in AI-Epistemic Scale (TAI-E) (2026) · cites this paper as Romeo and Conti 2026
Citations
Cites
The PRISMA 2020 statement: An updated guideline for reporting systematic reviews
Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., et al. (2021) · BMJ · doi:10.1136/bmj.n71 · Cited by 108,283
The PRISMA 2020 statement: an updated guideline for reporting systematic reviews
Page, M. J., McKenzie, J., Bossuyt, P., Boutron, I., Hoffmann, T., mulrow, C., Shamseer, L., Tetzlaff, J., Akl, E., Brennan, S. E., et al. (2020) · doi:10.31222/osf.io/v7gm2_v1 · Cited by 8,090
Trust in Automation: Designing for Appropriate Reliance
Lee, J. D., & See, K. A. (2004) · Human Factors: The Journal of the Human Factors and Ergonomics Society · doi:10.1518/hfes.46.1.50_30392 · Cited by 6,605 · cited 2×
“In the automation context, trust is defined as «the attitude that an (automated) agent will help achieve an individual’s goals in a situation characterized by uncertainty and vulnerability» (Lee and See 2004: p.51).”
The PRISMA 2020 statement: an updated guideline for reporting systematic reviews
Page, M. J., McKenzie, J., Bossuyt, P., Boutron, I., Hoffmann, T., mulrow, C., Shamseer, L., Tetzlaff, J., Akl, E., Brennan, S. E., et al. (2020) · doi:10.31222/osf.io/v7gm2 · Cited by 5,177
Rethlefsen, M. L., Kirtley, S., Waffenschmidt, S., Ayala, A. P., Moher, D., Page, M. J., Koffel, J. B., PRISMA-S Group, Blunt, H., Brigham, T., et al. (2021) · Systematic Reviews · doi:10.1186/s13643-020-01542-z · Cited by 3,714 · cited 2×
“In accordance with PRISMA-S guidelines (Rethlefsen et al. 2021), we also included records identified through personal files.”
Rethlefsen, M. L., Kirtley, S., Waffenschmidt, S., Ayala, A. P., Moher, D., Page, M. J., & Koffel, J. (2019) · doi:10.31219/osf.io/sfc38 · Cited by 3,696
What is AI Literacy? Competencies and Design Considerations
Long, D., & Magerko, B. (2020) · Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3313831.3376727 · Cited by 3,241 · cited 1×
“Key challenges lie in AI literacy, whose lack can limit user’s ability to effectively collaborate with systems (Long and Magerko 2020), and explanation design itself.”
Complacency and Bias in Human Use of Automation: An Attentional Integration
Parasuraman, R., & Manzey, D. H. (2010) · Human Factors: The Journal of the Human Factors and Ergonomics Society · doi:10.1177/0018720810376055 · Cited by 1,710 · cited 1×
“Some researchers emphasize differences in their cognitive and situational antecedents (Wickens et al. 2015), while others highlight substantial overlap, viewing both as manifestations of shared underlying attentional and decision-making mechanisms (Parasuraman and Manzey 2010).”
Automation bias: a systematic review of frequency, effect mediators, and mitigators
Goddard, K., Roudsari, A., & Wyatt, J. C. (2012) · Journal of the American Medical Informatics Association · doi:10.1136/amiajnl-2011-000089 · Cited by 1,162 · cited 7×
“Additionally, accountability, which has been proposed as an intervention to mitigate AB (Goddard et al. 2012), lacks empirical support.”
InfluentialBuçinca, Z., Malaya, M. B., & Gajos, K. Z. (2021) · arXiv · doi:10.48550/arxiv.2102.09692 · Cited by 1,124 · cited 12×
“In an effort to reduce over-reliance, Buçinca et al. (2021) adopted CFFs, a metacognitive debiasing technique aimed at promoting analytical thinking.”
Influential
Research works printed but not linked: 2
- Hemmer P, Schemmer M, Vössing M, Kühl N (2021) Human-AI complementarity in hybrid intelligence systems: a structured literature review. In: Pacific Asia conference on information systems, vol 78. ISBN 978-1-7336325-7-7
- Romeo G, Conti D (2024) Beyond automation: reshaping Human-artificial intelligence interaction. Sistemi Intelligenti XXVI(3):641648. https:// doi. org/ 10. 1422/ 115336 · doi:10.1422/115336
Cited by
Items 101–186 of 186
Kononykhina, O., Schierholz, M., & Kreuter, F. (2026) · Journal of Official Statistics · doi:10.1177/0282423x261455656
Wang, L. W., Ren, J., Sun, S., & Chen, Y. (2026) · Journal of theoretical and applied electronic commerce research · doi:10.3390/jtaer21070225
“On the one hand, automation bias reflects users’ excessive reliance on AI-generated outputs; on the other hand, algorithm aversion reflects users’ tendency to avoid or resist algorithmic recommendations after experiencing errors [32,33].”
Social calibration of sycophantic AI—Response
Cheng, M., Lee, C., Khadpe, P., & Jurafsky, D. (2026) · Science · doi:10.1126/science.aeh8601 · Cited by 1
Ogundokun, R. O., Bello, R.-W., Owolawi, P. A., Du, C., & Wyk, E. V. (2026) · International Conference on Human Centric Artificial Intelligence (ICHCAI) · doi:10.1109/ichcai70183.2026.11607648 · Cited by 1
“With that said, numerous surveys have found that the ease with which an explanation is understood depends heavily on the user's cognitive load [10].”
InfluentialBahadir, S., Syed, S. A., & D’Amico, R. S. (2026) · Neurosurgery · doi:10.1227/neu.0000000000004166
(Over)Reliance on Test Agents in AI-Assisted Software Testing
Enoiu, E. P. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-35506-5_25
Putri, A. S., Khairani, N. A., Sabrina, N., & Wijaya, M. I. (2026) · International Symposium on AI-Driven Engineering Systems (ISADES) · doi:10.1109/isades69945.2026.11608060
“Automation bias in the context of generative AI clarifies why AI users often accept biased AI output when coming to decisions [10][11].”
Clinical Artificial Intelligence Systems in Dentistry: Designing for Safe Human-AI Collaboration
Hegde, S., & Vasa, R. (2026) · Artificial intelligence · doi:10.5772/intechopen.1015701
An Integrated Framework for Human-Machine Trust Research in the Era of Large Language Models
朱, 孟., 陆, 文., & 柳, 林. (2026) · Global Journal of Media Studies · doi:10.26599/gjms.2026.9330013
Giacobbe, D. R., Lucis, R., Banerjee, S., Carra, G., Dexter, E., McFadden, B. R., & Rizzo, A. (2026) · Therapeutic Advances in Infectious Disease · doi:10.1177/20499361261467605 · Cited by 2
Network Orchestration Framework Design Using AI-Driven Automation and Cybersecurity
Annahdi, T., Alsumayt, A., & Alshammari, M. (2026) · Future Internet · doi:10.3390/fi18080394
Wang, H., Fu, F., Li, L., & Liu, M. (2026) · Frontiers in Psychology · doi:10.3389/fpsyg.2026.1861541 · Cited by 1
“This aligns with prior research showing that inappropriate trust in AI may encourage over-reliance and reduce individuals’ own monitoring and verification efforts (Hou et al., 2025; Romeo and Conti, 2026).”
Bunea, O.-I., Corboș, R.-A., & MIHAI, B. (2026) · Administrative Sciences · doi:10.3390/admsci16080360 · Cited by 1
“Greater domain knowledge may sometimes reduce reliance on automated systems because knowledgeable users have greater confidence in their own judgment (J. Lee & Moray, 1992; Romeo & Conti, 2026).”
Hendriyani, C., Sutarjo, S., Rahmayanti, I., & Nuryan, I. (2026) · F1000Research · doi:10.12688/f1000research.186714.1 · Cited by 1
Casallas, L. C. E., Chiappe, A., & Pérez, S. M. (2026) · Education Sciences · doi:10.3390/educsci16081209 · Cited by 2
Kopka, M., Aboueldahab, A., & Feufel, M. A. (2026) · Research Square · doi:10.21203/rs.3.rs-10348224/v1
Denecke, K., & Gabarrón, E. (2026) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2026.1785085 · Cited by 1
Ching-Tsung, J., Lu, W., & Sang, T. Q. (2026) · Research Square · doi:10.21203/rs.3.rs-10421674/v1
Does distrust priming reduce AI stereotypes? Testing applicability to AI acceptance
Dan, K., & Hoshino, T. (2026) · Computers in Human Behavior Reports · doi:10.1016/j.chbr.2026.101232
Rathomi, A., Marjuni, & Afifudin (2026) · SIWAYANG JOURNAL PUBLIKASI ILMIAH BIDANG PARIWISATA KEBUDAYAAN DAN ANTROPOLOGI · doi:10.47353/siwayang.v5i1.5233
Cognitive Readiness for Human-AI Collaboration
Tremblay, S., de Hemptinne, D., Teyssier‐Roberge, G., Gallant, A., Marois, A., & Lafond, D. (2026) · Human Factors The Journal of the Human Factors and Ergonomics Society · doi:10.1177/00187208261474345 · Cited by 5
pu, S., Liu, Y., & Wu, X. (2026) · Communications in computer and information science · doi:10.1007/978-981-92-0291-1_26
Siyu, J. (2026) · Advances in Psychology · doi:10.12677/ap.2026.168382
FarizRifqi., M., Mukti, I. Y., & Rijadi, S. C. R. (2026) · International Conference on Advancement in Data Science, E-learning and Information System (ICADEIS) · doi:10.1109/icadeis71120.2026.11644544
“…[11], [22], [34], [35] AI-Powered Valuation Analytics Engine [12], [13], [14], [15], [16], [17], [18], [21] XAI Interpretability Module [19], [20], [31], [36], [37] AI Decision Audit Logging Module [19], [20], [36], [37], [38] Appraisal Document Generator [12], [13], [18] Unified Customer Data…”
AI as a reflection partner in UX strategy: Presentation of an AI-supported UX strategy evaluator
Gronwald, F., & Rohles, B. (2026) · i-com · doi:10.1515/icom-2026-0049
“5,6 However, another observation was striking: several groups initially did not actively ask where statements came from or which assumptions they were based on.”
The efficiency–verification trade-off in large language model-assisted workplace risk assessment
Bosler, A., Koppel, T., & Reinhold, K. (2026) · Proceedings of the Estonian Academy of Sciences · doi:10.3176/proc.2026.3.10
EmoLoop: An adaptive VR biofeedback system for stress regulation
Lima, R., Chirico, A., Gaggioli, A., Gamboa, H. F. S., & Badia, S. B. I. (2026) · International Journal of Human-Computer Studies · doi:10.1016/j.ijhcs.2026.103916
Decoding user preferences: A study of explanations for algorithmic decision support in procurement
Ye, Y., Zheng, J., & Ma, L. (2026) · International Journal of Industrial Ergonomics · doi:10.1016/j.ergon.2026.104029
Dai, C., & Eichelberger, A. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-7902-9.ch003
Kocak, O., İdil, Ş., & Aydoğan, M. (2026) · Research Square · doi:10.21203/rs.3.rs-9615155/v1
Uddin, S. M. J., Jin, Z., & Chaudhary, N. (2026) · International Journal of Construction Management · doi:10.1080/15623599.2026.2717617
Arangi, D., Jhansi, S., Husain, M., Khare, Y., & Singh, B. (2026) · Journal of Progress in Artificial Intelligence · doi:10.68152/2326
Lemke, T., Marka, A. W., Prucker, P., Komenda, A., Weller, D., Mertens, C. J., Kim, S. H., Lammert, J., Huber, T., Graf, M. M., et al. (2026) · Academic Radiology · doi:10.1016/j.acra.2026.08.014 · Cited by 1
The Political Economy of AI Automation: Demystifying Its Impacts
Mah, V. A.-A., Khan, R., & Ramsurrun, N. (2026) · IGI Global eBooks · doi:10.4018/979-8-3373-9858-7.ch008
La información que nos da nuestro chatbot favorito no es neutral ni objetiva
Jiménez-Torras, D., & López, J. (2026) · doi:10.64628/aao.dm55d5jy4
Meng, K., Mahapatra, M. S., & Xiao, J. J. (2026) · Journal of Consumer Behaviour · doi:10.1002/cb.70238 · Cited by 1
Allgårdh, G., Helgesson, G., & Kihlbom, U. (2026) · BMC Medical Ethics · doi:10.1186/s12910-026-01572-7
“This risk may persist even in explainable systems, since explanations can increase perceived acceptability without necessarily improving decision accuracy and may, depending on their design, reinforce misplaced trust [39].”
Farnsworth, E. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-7709-4.ch006
Human-Centered Leadership in an AI-Empowered Workforce and the Synchronicity Triad
Geer, S. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-7709-4.ch005
Being More, Not Just Doing More: A Whole-Person, Human-Centric Perspective on AI
Dave, P., & Barresi, E. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-7709-4.ch001
Mızrak, F., & Karakaya, T. (2026) · Data · doi:10.3390/data11090215
“The pattern is consistent with research showing that effective human–AI collaboration depends on appropriate reliance rather than trust maximization [18,20,21].”
SSS Framework: A Think-First, AI-Second Model for AI-Assisted Writing.
Giray, L. (2026) · Annals of Biomedical Engineering · doi:10.1007/s10439-026-04358-5
Hammad, T. A., Rochon, J., Gofman, K., Trifirò, G., & Bate, A. (2026) · Drug Safety · doi:10.1007/s40264-026-01719-w
Hyytiäinen, T., & Luukka, P. (2026) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2026.2721889
Rafique, N., Mansoor, S., Zahid, M., & Irshad, A. (2026) · International Journal of Emerging Multidisciplinaries Social Science · doi:10.54938/ijemdss.2026.05.4.788
“…been concretely identified in the prior research that contrary to the adoption of earlier generations of technology, acceptance of AI technologies and capabilities, is challenging the traditional assumptions of certainty, user control, and transparency (Dwivedi, et al., 2021; Romeo & Conti, 2025).”
Sifakis, N., Sarti, S., Chachalis, A., Cholidis, D., Batistatos, M. C., Kourtis, M., Aryblia, M., & Arampatzis, G. (2026) · Sustainability · doi:10.3390/su18178858
Chao, L., & Ma, X. (2026) · Research Square · doi:10.21203/rs.3.rs-10644054/v1
Decision substitution and the reorganisation of judgement in AI-mediated society
Ibitoye, A., Ajenifuja, O. F., Kareem, R. S. A., Adeyemo, D. D., & Famutimi, R. F. (2026) · AI & Society · doi:10.1007/s00146-026-03349-3
Результаты пилотного исследования восприятия ответов ИИ
NOAKK, N. V., & Костина, Т. А. (2026) · Cifrovaâ èkonomika. · doi:10.34706/de-2026-04-06
Jurmu, M., Naumanen, M., Westerstrand, S., Lämsä, A., Myllymaa, K., & Ailisto, H. J. (2026) · Applied Artificial Intelligence · doi:10.1080/08839514.2026.2721301
Poyiadjis, K. (2026) · IGI Global eBooks · doi:10.4018/979-8-2600-0640-5.ch007
Opportunities, pitfalls, ethics, and clinical governance of AI in laboratory medicine
Lippi, G., Mattiuzzi, C., & Henry, B. M. (2026) · LabMed discovery. · doi:10.1016/j.lmd.2026.100145 · Cited by 1
Beyond Traditional Sales Training: An AI Driven Experiential Learning Framework
Singh, P. P., Saran, A., Khan, I., & Kalliny, M. (2026) · Marketing Education Review · doi:10.1080/10528008.2026.2729466
Popek, R. (2026) · Ethics & Behavior · doi:10.1080/10508422.2026.2729686
Pozzolini, M., Mongelli, G., Passarelli, M., & Rizzo, A. (2026) · European Conference on Cognitive Ergonomics · doi:10.1145/3822301.3822331
Altınışık, M. S. (2026) · Current Psychology · doi:10.1007/s12144-026-09999-2
“Automation bias primarily concerns uncritical acceptance of automated recommendations (Cummings, 2017 ; Romeo & Conti, 2025 ), cognitive offloading concerns the delegation of cognitive operations to external tools (Gieshoff & Heeb, 2023; Jose et al., 2025 ), and reduced self-efficacy concerns…”
Zhang, Y., Yang, R., & Jiang, Q. (2026) · Systems · doi:10.3390/systems14091150 · Cited by 1
Hoppe, T., Loibl, K., & Leuders, T. (2026) · Frontiers in Education · doi:10.3389/feduc.2026.1857661
Lüttke, J.-P., & Beutner, M. (2026) · Lecture notes in networks and systems · doi:10.1007/978-3-032-32729-1_18
Mızrak, F., Karakaya, T., & Durmaz, İ. B. V. (2026) · Journal of Intelligence · doi:10.3390/jintelligence14090223
“Fluid reasoning is expected to be particularly relevant because users must determine whether a conclusion follows from the available information rather than judge it by surface plausibility alone (Cheng et al., 2026; Romeo & Conti, 2026).”
Hart, G., Burrell, D. N., Valle, F., & Harper, M. (2026) · Revista Academiei Forţelor Terestre · doi:10.2478/raft-2026-0025
Bartz, A., Glaser, N., Bayer, O., Hallsson, L. R., Jaeger, V. K., Krumkamp, R., Minnerup, H., Pickford, R., Olsen, K. S., Wülk, N., et al. (2026) · Research Square · doi:10.21203/rs.3.rs-10974790/v1
Integrating AI Bots into Collaborative Requirements Decision-Making Through Temporal Graph Modeling
Mateos-Aparicio-Ruiz, I., Montealegre-Macias, P., López-Gómez, J. A., Déniz, Ó., & Bueno, G. (2026) · IEEE International Requirements Engineering Conference Workshops (REW) · doi:10.1109/rew72749.2026.00048
“When humans interact with AI-driven recommendations during requirements negotiation, they are highly susceptible to automation bias [4], inappropriate trust calibration [5] and the erosion of human oversight [6].”
Human-in-the-Loop Artificial Intelligence for Strategic Decision Support
Skander, S., & IBENRISSOUL, N. (2026) · International Conference on Connected Innovation and Technology (ICCITX) · doi:10.1109/iccitx70146.2026.11679931
Khullar, S., Andriani, A. D., Pamnani, H. K., & Saxena, R. K. (2026) · Advances in wireless technologies and telecommunication book series · doi:10.4018/979-8-2600-1533-9.ch010
Dóczi, Z., & Kristoffersson, M. (2026) · Computer law & security review · doi:10.1016/j.clsr.2026.106407
Formal Reasoning and Guidance Platform for Transparent Scaffolding for Multi-agent Debate
Harbar, Y., & Chudziak, J. A. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-36868-3_7
Who is Responsible for Tackling Professional Deskilling?
Kärki, K. (2026) · Human Affairs · doi:10.1515/humaff-2026-0080
“This is in line with the biases and heuristics literature showing that automation bias is linked with correlation bias (for discussion, see Romeo and Conti 2026), which means that the order of performing a decision and requesting a recommendation from arti fi cial intelligence a ff ects whether the…”
Development of a theoretical model for calibrating trust in AI to augment highways safety
Ivvala, P., Bayramova, A., Roberts, C. J., Edwards, D. J., & Rillie, I. (2026) · Smart and Sustainable Built Environment · doi:10.1108/sasbe-05-2026-0485
Niu, L., & Fu, S. (2026) · Journal of theoretical and applied electronic commerce research · doi:10.3390/jtaer21100338
Saini, J., Jassar, S., Adams, S., Alli, S. B., Colak, E., & Wilson, J. R. (2026) · Frontiers in Radiology · doi:10.3389/fradi.2026.1979076
“Confidence estimates and explanations may help users calibrate reliance, but poorly calibrated confidence, cognitively burdensome explanations, or overly simple rationales may instead reinforce misplaced trust (14, 15).”
Mal-Intelligence: how artificial intelligence produces a sustained illusion of cognitive competence
RAHHALI, M., & Aajly, A. (2026) · AI and Ethics · doi:10.1007/s43681-026-01398-4
“…than correcting [ 10 ],is structurally reinforced by sycophancy embedded in model training [ 4 ],is compounded by automation bias that resists educational correction [ 31, 34],and interacts with group-level normative pressure capable of scaling these effects beyond any single individual [ 6].”
InfluentialZhu, Z., & Li, Y. (2026) · Telematics and Informatics · doi:10.1016/j.tele.2026.102464 · Cited by 1
Groves, E., Baez-Lugo, S., Luthier, G., Rovescalli, A., Ribes, D., Henchoz, N., & Motta, M. (2026) · Nordic Conference on Human-Computer Interaction · doi:10.1145/3829807.3829871
Weng, Z., Mi, Y., Chen, X., & Li, Q. (2026) · Frontiers in Psychology · doi:10.3389/fpsyg.2026.1747252
Beyond the Algorithm: What Will It Take for Artificial Intelligence to Succeed in Indian Radiology?
Datta, S., Priyadarshini, P., & Rekhi, A. K. (2026) · Indian journal of radiology and imaging - new series/Indian journal of radiology and imaging/Indian Journal of Radiology & Imaging · doi:10.1055/s-0046-1829099
Abdulla, A. A. (2026) · International Journal of Law Management & Humanities · doi:10.63108/ijlmh.12875
Fneich, B., Ahadi, N., Turi, J. A., & Alkadhimi, F. (2026) · International Journal of Mental Health and Addiction · doi:10.1007/s11469-026-01724-1
Rafsanjani, M. A., Kurniawan, R. Y., Wahyudi, H. D., & Srisuk, P. (2026) · The Journal of Behavioral Science · doi:10.69523/tjbs.2026.298843 · Cited by 1
Palka, A., Kopyciński, W., Krawczyk, N., Wilczkowski, P., Krzysztofek, N., Kamela, M., Boszczyk, K., Grzybowska, M., Baran, A., & Krawczyk, O. (2026) · International Journal of Innovative Technologies in Social Science · doi:10.31435/ijitss.3(51).2026.6517
Agentic AI Trustworthiness Is a Full-Stack Problem: A Framework from Hardware to Governance
Farahani, B., Rouintan, M., Vahidinia, P., & Firouzi, F. (2026) · IEEE International Conference on Omni-layer Intelligent Systems (COINS) · doi:10.1109/coins69995.2026.11715577
“Reviewer training and rotation and “disagree” friction for critical approvals can help maintain active human judgment [30].”
Generative AI in supply chains: a qualitative study on decision-making and human-AI collaboration
Qarqash, Z., Putintseva, A., Gruchmann, T., & Ivanov, D. (2026) · Production Planning & Control · doi:10.1080/09537287.2026.2733698 · Cited by 3
Tran, D. L. A., & Dinh, L. N. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-2600-1488-2.ch012
Artificial Intelligence and the Radiographic Aging Clock
Katz, S. I., & Roshkovan, L. (2026) · CHEST Journal · doi:10.1016/j.chest.2026.07.5230
Using Artificial Intelligence to Help Human Commanders Fight Future Indo-Pacific Wars
Layton, P. (2026) · Journal of Asian Security and International Affairs · doi:10.1177/23477970261489106
Metadata
- Authors
- Romeo, G., & Conti, D.
- Affiliations
- University of Catania (IT)
- Corresponding author
- Romeo, Giuseppe
- Conti, Daniela
- Venue
- AI & SOCIETY
- Published
- 2025-07-03
- Volume
- 41
- Issue
- 1
- Pages
- 259-278
- ISSN
- 0951-5666 (ISSN-L)
- 1435-5655
- OpenAlex
- W4412013201
- Cited by
- 209
- References
- 60
- Citations by year
- 2026: 194 · 2025: 15
- FWCI
- 228.31
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- creativecommons.org/licenses/by/4.0 (tdm)
- creativecommons.org/licenses/by/4.0 (vor)
- Locations
- AI & Society (publishedVersion, cc-by)
- Topics
- Social Sciences › Human Factors and Ergonomics › Human-Automation Interaction and Safety
- Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Keywords
- automation bias, human-AI collaboration, explainable AI, trust in AI, AI literacy, explanation complexity, professional expertise, explanation design, user engagement, cognitive mechanisms
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-01, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-01 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.