Toward Resilient Human-AI Collaboration: A Lifecycle Taxonomy of Sociotechnical Risks and Cascading Failures

Isaac Kobby Anni; Md Foysal Ahmed; Md Main Uddin Rony

REVIEW 3 major objections 4 minor 107 references

Toward Resilient Human-AI Collaboration: A Lifecycle Taxonomy of Sociotechnical Risks and Cascading Failures

T0 review · 3 major / 4 minor · reviewed 2026-08-08 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Human-AI failures are cascades, not isolated bugs

desk verdict A creditable synthesis of human-AI collaboration risks whose cascade model is a hypothesis diagram, not a demonstrated result. read the letter →

arxiv 2608.05614 v1 pith:Y5HCSZLC submitted 2026-08-06 cs.HC

Md Foysal Ahmed , Isaac Kobby Anni , Md Main Uddin Rony This is my paper · ORCID

classification cs.HC

keywords human-AIcollaborationsociotechnicalrisksrisktaxonomycascadingfailurestrustcalibrationagenticAItechnostresslifecyclegovernance

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proposes that failures in human-AI collaboration are not isolated technical glitches but recurring sociotechnical breakdowns that cluster into six cross-domain risk types and cascade across a four-stage lifecycle. The four stages are task allocation, interaction, feedback, and adoption; the six clusters are trust miscalibration, cognitive burden, accountability gap, capability erosion, goal misalignment, and AI anxiety and technostress. A sympathetic reader would care because the taxonomy gives researchers and policymakers a shared language for comparing failures across healthcare, journalism, education, organizations, and defense, and because the cascade model predicts that fixing one risk in isolation can unintentionally worsen another. The paper is a synthesis and a conceptual model rather than an experiment, and it is offered as a foundation for empirical validation and lifecycle-oriented governance.

What carries the argument

The carrying object is a two-part framework: a four-stage lifecycle taxonomy (task allocation, interaction, feedback, adoption) and a conceptual interaction model, Figure 2. The taxonomy organizes the six risk clusters—trust miscalibration (over- or under-reliance on AI), cognitive burden (verification fatigue and overload), accountability gap (unclear responsibility and moral outsourcing), capability erosion (deskilling and loss of situational awareness), goal misalignment (AI optimizing the wrong objective), and AI anxiety and technostress (fear, exhaustion, burnout)—into recurring cross-domain categories. The interaction model links upstream drivers (task context, AI factors, human factors) to the six clusters and draws directional cascade arrows between clusters, such as cognitive burden pushing toward trust miscalibration and capability erosion feeding anxiety. Its work is to convert scattered evidence into a testable hypothesis about how localized failures become systemic.

What would settle it

A longitudinal field study in a single high-stakes domain could track the six clusters across the four lifecycle stages and test temporal precedence: if cognitive burden at the interaction stage does not predict increased trust miscalibration at the feedback stage, or if relieving cognitive burden leaves later trust outcomes unchanged, the directional cascade claim loses support.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that disparate domain-specific symptoms—clinical alert fatigue, journalistic accountability crises, deskilling in knowledge work, moral outsourcing in defense—are manifestations of the same six underlying risk clusters operating through shared mechanisms. The paper argues these clusters emerge from sociotechnical drivers such as task complexity, AI autonomy and opacity, and human expertise and AI literacy; propagate through a four-stage lifecycle; and interact through cascading pathways so that over-reliance, overload, and abdication feed one another. Many collaboration failures, it concludes, stem from interconnected sociotechnical dynamics rather than isolated technical deficiencies, which is why piecemeal interventions often produce unintended consequences such as the explainability paradox.

Load-bearing premise

The load-bearing premise is that the directional arrows in Figure 2 match real causal mechanisms—for example, cognitive burden driving trust miscalibration and capability erosion driving AI anxiety—and the paper itself states in Section 6.2 that the strength and direction of these relationships remain largely untested.

Editorial extensions

If this is right

If the taxonomy holds, an intervention that targets one risk cluster without considering the others—such as adding detailed AI explanations to build trust—can backfire by increasing cognitive burden and reinforcing automation bias.
The recurring clusters imply that risk evidence from one domain, such as healthcare alert fatigue, can inform governance in another, such as defense oversight, because the underlying failure mechanisms are shared.
Lifecycle governance should be organized around the four stages, with checks at task allocation, interaction, feedback, and adoption, rather than relying on one-time design fixes.
Multi-agent and agentic systems make the cascades harder to manage because humans become the coordination and verification layer, so failures propagate through inter-agent dependencies.
The framework sets up direct empirical targets: behavioral experiments and longitudinal studies can test whether the proposed risk clusters and cascade directions predict team performance and human well-being.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A testable extension the paper leaves open is a longitudinal instrument that measures all six clusters at each lifecycle stage to estimate whether cognitive burden at the interaction stage actually precedes trust miscalibration at the feedback stage.
The taxonomy could double as a coding scheme for incident reports and failure post-mortems; if the same six clusters recur in independent datasets, that would independently corroborate the unification claim.
The cascade model also predicts a practical ordering: interventions that jointly reduce cognitive burden and clarify accountability should outperform single-point fixes in randomized field trials, a comparison the paper does not run.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

You can treat this as a solid, clearly organized synthesis paper rather than a new empirical contribution. What's new here is not any single risk—automation bias, deskilling, alert fatigue, and accountability gaps all have deep literatures—but the packaging: six recurrent cross-domain risk clusters (Trust Miscalibration, Cognitive Burden, Accountability Gap, Capability Erosion, Goal Misalignment, AI Anxiety and Technostress) mapped onto a four-stage lifecycle (allocation, interaction, feedback, adoption) with a conceptual model of how those clusters interact and cascade. The authors do a good job showing that similar failure patterns recur across journalism, healthcare, defense, education, and knowledge work, and Table 1 is a reasonably useful map of that recurrence.

The paper is honest about its own status. Section 6.2 and the Limitations section explicitly say that the causal arrows in Figure 2 are a conceptual proposal rather than an empirically validated account. That is the right framing.

The main soft spot is exactly where the stress-test note lands: the cascade model's arrows are at risk of being partly definitional. For instance, automation bias appears both as a constituent symptom of Trust Miscalibration (§4.1) and as a downstream consequence of Cognitive Burden (§4.2, §5.1.1); learned helplessness shows up under both Adoption risks (§2.4) and Accountability Gap (§4.3). When symptoms belong to multiple clusters and also serve as cross-cluster outcomes, the model can confirm an arrow by construction. This overlap is not fatal for a qualitative taxonomy—real-world risks do blur—but it does mean Figure 2 is a hypothesis diagram, not a demonstrated model. The paper would be stronger with explicit rules for assigning symptoms to clusters, or at least a candid discussion of how overlap affects the cascade claims.

A second, minor issue: the domain selection and cluster mapping in Table 1 are asserted rather than derived from a systematic literature protocol. That is normal for a narrative review, and the authors do not overclaim, so I would call it minor.

Who should read this? Researchers new to human-AI collaboration, or people wanting a concise map of the risk landscape and a springboard for governance work. It will not change the research agenda of specialists. It deserves peer review—not because it breaks new ground, but because it is a competent, potentially useful synthesis that should be checked by experts. I would accept it with encouragement to tighten the cluster definitions and soften the causal language to match the evidence.

Referee Report

3 major / 4 minor

Summary. This paper presents a lifecycle-oriented synthesis of sociotechnical risks in human-AI collaboration. It organizes risks into four lifecycle stages (task allocation, interaction, feedback, adoption) and proposes six cross-domain risk clusters: Trust Miscalibration, Cognitive Burden, Accountability Gap, Capability Erosion, Goal Misalignment, and AI Anxiety and Technostress. The authors also introduce a conceptual interaction model (Figure 2) in which sociotechnical drivers activate these risk clusters and the clusters interact through cascading pathways that ultimately affect team performance and human well-being. The paper is a qualitative, literature-based synthesis; it does not report new empirical data and explicitly acknowledges in Section 6.2 and Section 7 that the proposed causal relationships remain untested.

Significance. If the taxonomy and cascade model are accepted as a valid organizing framework, the paper would provide a useful cross-domain vocabulary for researchers and practitioners working on human-AI collaboration. The synthesis of fragmented literatures and the emphasis on interconnected, lifecycle-dependent failures are valuable contributions, and the authors are candid about the theoretical status of the interaction model. The paper does not overclaim empirical support. However, the central claim that collaboration failures stem from interconnected cascades rather than isolated causes depends on the directional arrows in Figure 2, and the current manuscript does not yet provide a way to test or even uniquely instantiate those arrows because the risk clusters are not defined with mutually exclusive symptoms. This is a hypothesis-generating framework, and its value will depend on whether the authors can sharpen the operational definitions and empirical agenda.

major comments (3)

[§4.1, §4.2, §5.1.1, Figure 2] The cluster definitions overlap, which makes the cascade arrows in Figure 2 partly definitional rather than empirical. Automation bias is described as a constituent symptom of Trust Miscalibration in §4.1, but §4.2 and §5.1.1 also treat automation bias as a consequence of Cognitive Burden; similarly, learned helplessness appears under Accountability Gap in §4.3 and under Adoption risks in §2.4. When a symptom is both a component of one cluster and an outcome of another, the model can 'confirm' a cascade by construction. To make the central claims falsifiable, the authors should either define the six clusters with mutually exclusive symptom sets or explicitly recast Figure 2's arrows as testable hypotheses with operationalized measures and predicted directions.
[§5.1.1 and §6.2] The directional arrows in Figure 2 are asserted from selected citations, yet §6.2 concedes that 'the strength and directionality of the proposed risk relationships remain largely untested.' This is load-bearing because the paper's central message—that failures are interconnected and that piecemeal interventions are therefore inadequate—depends on these arrows. The manuscript should either soften the conclusion to a hypothesis-generation contribution or include a concrete empirical validation protocol (e.g., specific system dynamics models, longitudinal panel designs, or cross-lagged analyses) as part of the paper's contribution, rather than only as future work.
[Table 1] The assignment of references to risk clusters in Table 1 is presented as an evidentiary mapping, but no systematic selection or coding protocol is provided. For instance, reference [22] appears under both Cognitive Burden and Goal Misalignment, and it is not explained why a given symptom is placed in one cluster rather than another. Without a transparent procedure or an explicit statement that this is an interpretive synthesis rather than a systematic evidence map, the cluster assignments risk reflecting author selection rather than the cited evidence. The authors should either document the inclusion and clustering criteria or recharacterize Table 1 as an illustrative, non-systematic mapping.

minor comments (4)

[Figure 1 caption] The caption contains a typo: 'Collabroation' should be 'Collaboration', and the stage list 'Feedback adoption' should probably read 'Feedback, Adoption' with a comma.
[§2.4] The sentence 'Workers’ routine is outsourcing complex work to machines' is grammatically unclear; consider revising to something like 'Workers routinely outsource complex work to machines, losing the practice of maintaining their own cognitive edge.'
[§6.1] The last sentence of Section 6.1 ends without a period: '... in complex collaborative settings' is missing terminal punctuation.
[§7] The limitations section is commendably honest, but it would be more useful if it explicitly noted that the overlap among clusters (automation bias, learned helplessness) is itself a limitation affecting the testability of the model, not only the need for empirical validation.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper is an explicitly hypothesis-generating literature synthesis, and its causal cascade arrows are not derived from its own definitions.

full rationale

This manuscript contains no equations, fitted parameters, numerical predictions, or formal derivation chain, so the standard circularity failure modes (self-definitional reduction, fitted input called prediction, uniqueness imported from authors, ansatz smuggled via citation, renaming known result) do not apply. The six risk clusters are presented as an interpretive abstraction from diverse domain literatures (Sections 3 and 4), and the Figure 2 interaction model is explicitly described as a 'conceptual interaction model' whose 'strength and directionality ... remain largely untested' (Section 6.2); Section 7 likewise states that 'our systemic interaction framework is inherently theoretical, and its specific causal pathways and cascading effects require direct empirical validation.' The claim that cognitive burden can lead to trust miscalibration is supported by external empirical and review citations ([60], [71], [78]), not by the taxonomy's own definitions. The overlap of automation bias under Trust Miscalibration and Cognitive Burden is a construct-clarity or mutual-exclusivity concern, not a circular reduction, because the paper neither defines the clusters as mutually exclusive nor derives the cascade arrows from cluster membership alone. There are no self-citations that carry load-bearing weight: all references are to external work. The central message is a synthesis and a research agenda, with its causal arrows candidly marked as hypotheses rather than demonstrated results, so no circularity is present.

Assumptions & free parameters 0 free parameters · 4 assumptions · 2 invented entities

This paper makes no quantitative claims, so there are no fitted parameters. Its central contribution depends on interpretive assumptions: the lifecycle stages, the representativeness of the five domains, the mapping of symptoms to clusters in Table 1, and the causal cascade arrows in Figure 2. These are reasonable as a literature-synthesis scaffold but are not independently evidenced.

assumptions (4)

domain assumption The four lifecycle stages, task allocation, interaction, feedback, and adoption, are a valid sequential decomposition of human-AI collaboration.
Introduced in Section 2; if the stages overlap or are non-exhaustive, the temporal narrative of early risks shaping later outcomes is weakened.
domain assumption The five domains chosen, media, organization, education, healthcare, and defense, are representative sentinel contexts for cross-domain generalization.
Section 3 frames these as sentinel contexts and Table 1 maps clusters across them. Section 7 acknowledges that the underlying literature is predominantly Western and high-resource, which limits generalizability.
ad hoc to paper The cluster assignments in Table 1 reflect the cited evidence rather than author selection.
No systematic review protocol, intercoder reliability, or corpus selection criteria are reported in Section 4; the mapping from domain symptoms to clusters is asserted.
domain assumption The six risk clusters have stable meaning across domains such that the same label, for example cognitive burden, captures the same underlying vulnerability in each domain.
Section 4 abstracts domain symptoms into unified clusters; if the constructs are domain-specific in kind rather than in degree, the cross-domain claims would be weakened.

invented entities (2)

Six cross-domain risk clusters as reified categories
purpose: Classify diverse human-AI collaboration failures under six shared vulnerability types.
The clusters are introduced as analytic categories in Section 4, not derived from quantitative data or machine-checked inference; Section 6.2 calls for empirical validation.
Conceptual interaction model and cascade arrows
purpose: Explain how risk clusters trigger one another and affect team performance and human well-being.
Figure 2 is a qualitative diagram; the directional causal arrows are hypotheses. Section 7 states that the framework is inherently theoretical.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Toward Resilient Human-AI Collaboration: A Lifecycle Taxonomy of Sociotechnical Risks and Cascading Failures." pith.science (2026). https://pith.science/paper/Y5HCSZLC

@misc{pith2026260805614,
  author       = {Pith},
  title        = {Pith review of: Toward Resilient Human-AI Collaboration: A Lifecycle Taxonomy of Sociotechnical Risks and Cascading Failures},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Y5HCSZLC}},
  note         = {Machine review of arXiv:2608.05614}
}

read the original abstract

As AI systems become increasingly integrated into consequential domains such as healthcare, journalism, education, scientific research, organizational decision-making, and defense, effective human-AI collaboration has emerged as a critical challenge. However, the sociotechnical risks that undermine collaboration are often studied in isolation, obscuring the recurring failure mechanisms that cut across domains. This paper presents a lifecycle-oriented synthesis of human-AI collaboration risks spanning four stages: task allocation, interaction, feedback, and adoption. Drawing on evidence from diverse application domains, we identify six recurring cross-domain risk clusters: Trust Miscalibration, Cognitive Burden, Accountability Gap, Capability Erosion, Goal Misalignment, and AI Anxiety and Technostress. We further propose a conceptual interaction model that illustrates how these risks emerge from sociotechnical drivers, interact through cascading pathways, and ultimately affect team performance and human well-being. Our analysis shows that many collaboration failures stem not from isolated technical deficiencies but from interconnected sociotechnical dynamics, helping explain why piecemeal interventions frequently create unintended consequences. By synthesizing fragmented literature into a unified framework, this work provides a foundation for future empirical research, lifecycle-oriented governance, and the design of more resilient, trustworthy, and human-centered human-AI collaboration systems.

Figures

Figures reproduced from arXiv: 2608.05614 by the authors.

**Figure 2.** Conceptual interaction model of human-AI collab [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

107 extracted references · 69 canonical work pages

[22]

Robert Eccles. 2025. Hybrid Intelligence Teams: A Theoretical Framework for Human-AI Collaboration in Knowledge Work. A vailable at SSRN 5792345 (2025)

2025
[1]

Habeeb Abdulrauf, Abdulmalik Adetola Lawal, Amarachi Nina Uma Mba, Com- fort Ademola, Zaynab B Yusuf, Shalewa Babatayo, and Idris Ayinde. 2025. Artifi- cial Intelligence in Journalism: A Narrative Review of Opportunities, Challenges, Ethical Tensions, and Human-Machine Collaboration. American Journal of Arts and Human Science 4, 4 (2025), 5–18

2025
[2]

Manal Ahdadou, Abdellah Aajly, and Mohamed Tahrouch. 2024. Unlocking the potential of augmented intelligence: a discussion on its role in boardroom decision-making. International Journal of Disclosure and Governance 21, 3 (2024), 433–446

2024
[3]

NIST AI. 2023. Artificial intelligence risk management framework (AI RMF 1.0). URL: https://nvlpubs. nist. gov/nistpubs/ai/nist. ai (2023), 100–1

2023
[4]

Mohamed Adlan Ait Ameur, Erfu Yang, and Scott Brady. 2024. Advancing healthcare through intelligent human-robot collaboration: overview, challenges and opportunities. In 2024 29th International Conference on Automation and Computing (ICAC). IEEE, 1–6

2024
[5]

Saara Ala-Luopa, Sami Koivunen, Thomas Olsson, and Kaisa Väänänen. 2024. Considerations on human-AI collaboration in knowledge work–recruitment experts’ needs and expectations. (2024)

2024
[6]

Ying Bao, Xusen Cheng, Triparna De Vreede, and Gert-Jan De Vreede. 2021. Investigating the relationship between AI and trust in human-AI collaboration. (2021)

2021
[7]

Giulia Bassi, Valeria Orso, Silvia Salcuni, and Luciano Gamberini. 2025. Un- derstanding Workers’ Well-Being and Cognitive Load in Human-Cobot Col- laboration: Systematic Review. Journal of Medical Internet Research 27 (2025), e75658

2025

Show all 107 references

[8]

Michał Boni. 2021. The ethical dimension of human–artificial intelligence collaboration. European View 20, 2 (2021), 182–190

2021
[9]

Tamer Boyacı, Caner Canyakmaz, and Francis De Véricourt. 2024. Human and machine: The impact of machine input on decision making under cognitive limitations. Management Science 70, 2 (2024), 1258–1275

2024
[10]

Zana Buçinca, Maja Barbara Malaya, and Krzysztof Z Gajos. 2021. To trust or to think: cognitive forcing functions can reduce overreliance on AI in AI-assisted decision-making. Proceedings of the ACM on Human-computer Interaction 5, CSCW1 (2021), 1–21

2021
[11]

Ricardo Buettner. 2013. Cognitive workload of humans using artificial intelli- gence systems: towards objective measurement applying eye-tracking technol- ogy. In Annual conference on artificial intelligence . Springer, 37–48

2013
[12]

Katharina Buschmeyer, Sarah Hatfield, and Julie Zenner. 2023. Psychological as- sessment of AI-based decision support systems: tool development and expected benefits. Frontiers in Artificial Intelligence 6 (2023), 1249322

2023
[13]

Ángel Alexander Cabrera, Adam Perer, and Jason I Hong. 2023. Improving human-AI collaboration with descriptions of AI behavior. Proceedings of the ACM on Human-Computer Interaction 7, CSCW1 (2023), 1–21

2023
[14]

Shiye Cao, Catalina Gomez, and Chien-Ming Huang. 2023. How time pres- sure in different phases of decision-making influences human-AI collaboration. Proceedings of the ACM on Human-computer Interaction 7, CSCW2 (2023), 1–26

2023
[15]

Shiye Cao, Anqi Liu, and Chien-Ming Huang. 2024. Designing for appropriate reliance: The roles of AI uncertainty presentation, initial user decision, and user demographics in AI-assisted decision-making. Proceedings of the ACM on Human-Computer Interaction 8, CSCW1 (2024), 1–32

2024
[16]

Na Chen and Xinyue Zhang. 2025. When misunderstanding meets artificial intelligence: the critical role of trust in human–AI and human–human team communication and performance. Frontiers in Psychology 16 (2025), 1637339

2025
[17]

Zhisheng Chen. 2023. Collaboration among recruiters and artificial intelligence: removing human prejudices in employment. Cognition, Technology & Work 25, 1 (2023), 135–149

2023
[18]

Bianca Ifeoma Chigbu and Sicelo Leonard Makapela. 2025. AI in education, sustainability, and the future of work: An integrative review of industry 5.0, education 5.0, and work 5.0. Journal of Open Innovation: Technology, Market, and Complexity (2025), 100645

2025
[19]

Belle Dang, Luna Huynh, Faaiz Gul, Carolyn Rosé, Sanna Järvelä, and Andy Nguyen. 2025. Human–AI collaborative learning in mixed reality: Examining the cognitive and socio-emotional interactions. British Journal of Educational Technology 56, 5 (2025), 2078–2101

2025
[20]

Lilian Tai Do Khac and Michael Leyer. 2025. Towards an integrative model of organizational human-AI collaboration: A semi-systematic review of the current state of the art. Technology in Society (2025), 103064

2025
[21]

Alpana Dubey, Kumar Abhinav, Sakshi Jain, Veenu Arora, and Asha Puttaveer- ana. 2020. HACO: a framework for developing human-AI teaming. InProceedings of the 13th innovations in software engineering conference (formerly known as india software engineering conference) . 1–9. Md...

2020
[23]

Mica R Endsley and Esin O Kiris. 1995. The out-of-the-loop performance problem and level of control in automation. Human factors 37, 2 (1995), 381– 394

1995
[24]

Filipa Ferrada and Luis M Camarinha-Matos. 2024. Emotions in human-AI collaboration. In Working Conference on Virtual Enterprises . Springer, 101–117

2024
[25]

George Fragiadakis, Christos Diou, George Kousiouris, and Mara Nikolaidou
[26]

Ismael T Freire, Oscar Guerrero-Rosado, Adrián F Amil, and Paul FMJ Verschure
[27]

Andreas Fügener, Jörn Grahl, Alok Gupta, and Wolfgang Ketter. 2022. Cognitive challenges in human–artificial intelligence collaboration: Investigating the path toward productive delegation. Information systems research 33, 2 (2022), 678–696

2022
[28]

Frontiers in Robotics and AI 11 (2024), 1248646

Socially adaptive cognitive architecture for human-robot collaboration in industrial settings. Frontiers in Robotics and AI 11 (2024), 1248646

2024
[29]

Ali Gebeşçe, Muge Kural, Tilek Chubakov, and Gözde Gül Şahin. 2025. Quantify- ing Divergence for Human-AI Collaboration and Cognitive Trust. InProceedings of the Extended Abstracts of the CHI Conference on Human Factors in Computing Systems. 1–10

2025
[30]

Alessandro Gambetti, Qiwei Han, Hong Shen, and Claudia Soares. 2025. A survey on human-centered evaluation of explainable AI methods in clinical decision support systems. arXiv preprint arXiv:2502.09849 (2025)

2025
[31]

Hao Guo, Bangan Wu, Qi Li, Zhen Ding, Feng Jiang, and Chunzhi Yi. 2024. Im- pact of Cognitive Load on Human Trust in Hybrid Human-Robot Collaboration. arXiv preprint arXiv:2412.20654 (2024)

2024 arXiv
[32]

Meike Grimme and Christian Zabel. 2025. AI in the newsroom: A collective case study about newsworker-AI collaboration in the German newspaper industry. Journal of Media Business Studies 22, 2 (2025), 118–142

2025
[33]

Lewis Hammond, Alan Chan, Jesse Clifton, Jason Hoelscher-Obermaier, Akbir Khan, Euan McLean, Chandler Smith, Wolfram Barfuss, Jakob Foerster, Tomáš Gavenčiak, et al . 2025. Multi-agent risks from advanced ai. arXiv preprint arXiv:2502.14143 (2025)

2025 arXiv
[34]

ASHLESHA GUPTA, ANUPAM MUND, SHWETA ROY, PUNEET GARG, and DINESH KUMAR YADAV. 2025. TRUST IN AI SYSTEMS: A SOCIAL- PSYCHOLOGICAL INVESTIGATION OF HUMAN–AI COLLABORATION. TPM–Testing, Psychometrics, Methodology in Applied Psychology 32, S7 (2025): Posted 10 October (2025), 428–446

2025
[35]

Savindu Herath Pathirannehelage, Yash Raj Shrestha, and Georg von Krogh
[36]

Steffen Heim and Sylvia Chan-Olmsted. 2023. Consumer trust in AI–human news collaborative continuum: Preferences and influencing factors by news production phases. Journalism and Media 4, 3 (2023), 946–965

2023
[37]

Joshua Holstein and Gerhard Satzger. 2025. Development of Mental Mod- els in Human-AI Collaboration: A Conceptual Framework. arXiv preprint arXiv:2510.08104 (2025)

2025
[38]

Jinwei Hu, Yi Dong, Shuang Ao, Zhuoyun Li, Boxuan Wang, Lokesh Singh, Guangliang Cheng, Sarvapali D Ramchurn, and Xiaowei Huang. 2025. Stop Reducing Responsibility in LLM-Powered Multi-Agent Systems to Local Align- ment. arXiv preprint arXiv:2510.14008 (2025)

2025
[39]

Jimpei Hitsuwari, Yoshiyuki Ueda, Woojin Yun, and Michio Nomura. 2023. Does human–AI collaboration lead to more creative art? Aesthetic evaluation of human-made and AI-generated haiku poetry. Computers in Human Behavior 139 (2023), 107502

2023
[40]

Kashif Imteyaz, Michael Muller, Claudia Flores-Saviaga, and Saiph Savage. 2026. Co-Designing Collaborative Generative AI Tools for Freelancers. In Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems . 1–21

2026
[41]

Mohammad Hossein Jarrahi. 2018. Artificial intelligence and the future of work: Human-AI symbiosis in organizational decision making. Business horizons 61, 4 (2018), 577–586

2018
[42]

Shunsen Huang, Xiaoxiong Lai, Li Ke, Yajun Li, Huanlei Wang, Xinmei Zhao, Xinran Dai, and Yun Wang. 2024. AI technology panic—is AI dependence bad for mental health? A cross-lagged panel model and the mediating roles of motivations for AI use among adolescents. Psychology Rese...

2024
[43]

Aaron W Johnson, Kevin R Duda, Thomas B Sheridan, and Charles M Oman
[44]

Elizabeth A Johnson, Katherine M Dudding, and Jane M Carrington. 2024. When to err is inhuman: An examination of the influence of artificial intelligence- driven nursing care on patient safety. Nursing Inquiry 31, 1 (2024), e12583

2024
[45]

Mohammad Hossein Jarrahi, David Askay, Ali Eshraghi, and Preston Smith
[46]

Sue E Kase, Chou P Hung, Tomer Krayzman, James Z Hare, B Christopher Rinderspacher, and Simon M Su. 2022. The future of collaborative human- artificial intelligence decision-making for mission planning. Frontiers in Psy- chology 13 (2022), 850628

2022
[47]

Byung-Jik Kim and Julak Lee. 2024. The mental health implications of artificial intelligence adoption: the crucial role of self-efficacy. Humanities and Social Sciences Communications 11, 1 (2024), 1561

2024
[48]

Justin Kim, Taylor Davis, and Lingzi Hong. 2022. Augmented intelligence: enhancing human decision making. InBridging Human Intelligence and Artificial Intelligence. Springer, 151–170

2022
[49]

Vegard Kolbjørnsrud. 2024. Designing the intelligent organization: Six principles for human-AI collaboration. California Management Review 66, 2 (2024), 44–64

2024
[50]

James Johnson. 2022. The AI commander problem: Ethical, political, and psy- chological dilemmas of human-machine interactions in AI-enabled warfare. Journal of Military Ethics 21, 3-4 (2022), 246–271

2022
[51]

Yi Lai, Atreyi Kankanhalli, and Desmond Ong. 2021. Human-AI collaboration in healthcare: A review and research agenda. (2021)

2021
[52]

Matti Laukkarinen. 2025. Working with and around artificial intelligence: AI crafting and human-AI collaboration in recruitment. International Journal of Human–Computer Interaction 41, 21 (2025), 13804–13816

2025
[53]

Jiayao Li, Alexis Evain, Clara Le Duff, Tianyuan Zhang, David Panzoli, Audrey Fertier, Benoit Montreuil, and Frederick Benaben. 2026. Unravelling Cascading Phenomena in Sociotechnical Systems: A Systematic Review and Modelling Framework. Systems Research and Behavioral Science (2026)

2026
[54]

Jingshu Li, Yitian Yang, Renwen Zhang, Q Vera Liao, Tianqi Song, Zhengtao Xu, and Yi-chieh Lee. 2024. Understanding the Effects of Miscalibrated AI Confidence on User Trust, Reliance, and Decision Efficacy. arXiv preprint arXiv:2402.07632 (2024)

2024
[55]

Tim Robin Kosack and Thomas Krüger. 2025. Human-AI Collaboration in Cogni- tive Warfare: Enhancing Resilience Against Information Operations.Meaningful Human Control in Information Warfare (2025)

2025
[56]

Bowen Lou, Tian Lu, TS Raghu, and Yingjie Zhang. 2025. Unraveling human-AI teaming: a review and outlook. arXiv preprint arXiv:2504.05755 (2025)

2025 arXiv
[57]

Are you really sure?

Shuai Ma, Xinru Wang, Ying Lei, Chuhan Shi, Ming Yin, and Xiaojuan Ma. 2024. “Are you really sure?” Understanding the effects of human self-confidence cali- bration in AI-assisted decision making. InProceedings of the 2024 CHI Conference on Human Factors in Computing Systems . 1–20

2024
[58]

Neel Niladri Majumder and Babatope O Adebiyi. 2026. Human-Centered AI in Healthcare. In Handbook of Human-Centered Artificial Intelligence . Springer, 1–55

2026
[59]

Dayan Sanjeewa Munasinghe. 2024. Risk Propagation Simulator for Assessing Urban Risk. Ph. D. Dissertation. University of Salford (United Kingdom)

2024
[60]

Fanjue Liu, Mo Chen, and Seungahn Nah. 2026. Who writes the news matters: the role of social trust in shaping credibility across AI, human and human–AI collaboration. Online Information Review (2026), 1–19

2026
[61]

Adetoyese Omoseebi, Jack Enderson, and Alex Corner. 2025. Human-AI Collab- oration in Conflict Scenarios. (2025)

2025
[62]

Thomas O’neill, Nathan McNeese, Amy Barron, and Beau Schelble. 2022. Human–autonomy teaming: A review and analysis of the empirical literature. Human factors 64, 5 (2022), 904–938

2022
[63]

Antonio Padovano and Martina Cardamone. 2024. Towards human-AI collabo- ration in the competency-based curriculum development process: The case of industrial engineering and management education. Computers and Education: Artificial Intelligence 7 (2024), 100256

2024
[64]

Sun Young Park, Pei-Yi Kuo, Andrea Barbarin, Elizabeth Kaziunas, Astrid Chow, Karandeep Singh, Lauren Wilcox, and Walter S Lasecki. 2019. Identifying challenges and opportunities in human-AI collaboration in healthcare. In Com- panion Publication of the 2019 Conference on Comp...

2019
[65]

Bheema Shanker Neyigapula. 2023. Human-AI Collaborative Decision-making: A Cognitive Ergonomics Approach. (2023)

2023
[66]

Peng Peng, Zuocai Wang, Peng Cui, Xiaokang Hu, Junfeng Yao, and Sainan Lyu
[67]

Maite Puerta-Beldarrain, Oihane Gómez-Carmona, Rubén Sánchez-Corcuera, Diego Casado-Mansilla, Diego Lopez-de Ipina, and Liming Chen. 2025. A multifaceted vision of the human-AI collaboration: A comprehensive review. IEEE Access 13 (2025), 29375–29405

2025
[68]

Chandan K Reddy and Parshin Shojaee. 2025. Towards scientific discovery with generative ai: Progress, opportunities, and challenges. In Proceedings of Toward Resilient Human–AI Collaboration: A Lifecycle Taxonomy of Sociotechnical Risks and Cascading Failures the AAAI conferen...

2025
[69]

Carlo Reverberi, Tommaso Rigon, Aldo Solari, Cesare Hassan, Paolo Cherubini, and Andrea Cherubini. 2022. Experimental evidence of effective human–AI collaboration in medical decision-making. Scientific reports 12, 1 (2022), 14952

2022
[70]

Bhavik N Patel, Louis Rosenberg, Gregg Willcox, David Baltaxe, Mimi Lyons, Jeremy Irvin, Pranav Rajpurkar, Timothy Amrhein, Rajan Gupta, Safwan Halabi, et al. 2019. Human–machine partnership with artificial intelligence for chest radiograph diagnosis. NPJ digital medicine 2, 1...

2019
[71]

Giuseppe Romeo and Daniela Conti. 2026. Exploring automation bias in human– AI collaboration: a review and implications for explainable AI. Ai & Society 41, 1 (2026), 259–278

2026
[72]

Frontiers in Public Health 13 (2025), 1686346

Data-driven hierarchical causal modeling of risk propagation in bridge operations: evidence from 132 accidents in China. Frontiers in Public Health 13 (2025), 1686346

2025
[73]

Ranjan Sapkota, Konstantinos I Roumeliotis, and Manoj Karkee. 2025. Ai agents vs. agentic ai: A conceptual taxonomy, applications and challenges. Information Fusion (2025), 103599

2025
[74]

Emma Schleiger, Claire Mason, Claire Naughtin, Andrew Reeson, and Cecile Paris. 2024. Collaborative intelligence: A scoping review of current applications. Applied Artificial Intelligence 38, 1 (2024), 2327890

2024
[75]

Johanna Schmidhuber, Stephan Schlögl, and Christian Ploder. 2021. Cognitive load and productivity implications in human-chatbot interaction. In 2021 IEEE 2nd international conference on human-machine systems (ICHMS) . IEEE, 1–6

2021
[76]

Jeba Rezwana and Mary Lou Maher. 2022. Identifying ethical issues in ai partners in human-ai co-creation. arXiv preprint arXiv:2204.07644 (2022)

2022 arXiv
[77]

Scott Spillias, Paris Tuohy, Matthew Andreotta, Ruby Annand-Jones, Fabio Boschetti, Christopher Cvitanovic, Joseph Duggan, Elisabeth A Fulton, Denis B Karcher, Cecile Paris, et al. 2024. Human-AI collaboration to identify literature for evidence synthesis. Cell Reports Sustain...

2024
[78]

Budi Santoso and Rina Wijayanti. 2024. Human-AI Collaboration in Creative Industries: Workflows in Media Production and Community-Driven Platforms. Transactions on Artificial Intelligence, Machine Learning, and Cognitive Systems 9, 11 (2024), 11–26

2024
[79]

Mark Steyvers and Lukas Mayer. 2025. When not to help: planning for lasting human-AI collaboration. arXiv preprint arXiv:2508.01837 (2025)

2025 arXiv
[80]

Mira Suryani, Harry Budi Santoso, Martin Schrepp, Rizal Fathoni Aji, Setiawan Hadi, Dana Indra Sensuse, Ryan Randy Suryono, et al. 2024. Role, methodol- ogy, and measurement of cognitive load in computer science and information systems research. IEEE Access 12 (2024), 190007–190024

2024
[81]

Mirac Suzgun, Tayfun Gur, Federico Bianchi, Daniel E Ho, Thomas Icard, Dan Jurafsky, and James Zou. 2025. Language models cannot reliably distinguish belief from knowledge and fact. Nature Machine Intelligence (2025), 1–11

2025
[82]

Chuhan Shi, Xiaoquan Ren, Yifang Wang, Junze Li, Yushi Sun, Yawen Luo, and Rui Sheng. 2026. A Survey of Human-AI Collaboration for Scientific Discovery. Preprints (2026)

2026
[83]

Richard Tomsett, Alun Preece, Dave Braines, Federico Cerutti, Supriyo Chakraborty, Mani Srivastava, Gavin Pearson, and Lance Kaplan. 2020. Rapid trust calibration through interpretable and uncertainty-aware AI. Patterns 1, 4 (2020)

2020
[84]

Mark Steyvers and Aakriti Kumar. 2024. Three challenges for AI-assisted decision-making. Perspectives on Psychological Science 19, 5 (2024), 722–734

2024
[85]

Michelle Vaccaro, Abdullah Almaatouq, and Thomas Malone. 2024. When combinations of humans and AI are useful: A systematic review and meta- analysis. Nature Human Behaviour 8, 12 (2024), 2293–2303

2024
[86]

Karel Van Den Bosch, Adelbert Bronkhorst, et al. 2018. Human-AI cooperation to benefit military decision making. In Nato, Vol. 3. 2

2018
[87]

Karel Van Den Bosch, Tjeerd Schoonderwoerd, Romy Blankendaal, and Mark Neerincx. 2019. Six challenges for human-AI Co-learning. In International Conference on Human-Computer Interaction . Springer, 572–589

2019
[88]

Arthur Thuy and Dries F Benoit. 2024. Explainability through uncertainty: Trustworthy decision-making with neural networks. European Journal of Oper- ational Research 317, 2 (2024), 330–340

2024
[89]

Dakuo Wang, Justin D Weisz, Michael Muller, Parikshit Ram, Werner Geyer, Casey Dugan, Yla Tausczik, Horst Samulowitz, and Alexander Gray. 2019. Human-AI collaboration in data science: Exploring data scientists’ percep- tions of automated AI. Proceedings of the ACM on human-com...

2019
[90]

Usman Ahmad Usmani, Ari Happonen, and Junzo Watada. 2023. Human- centered artificial intelligence: Designing for user empowerment and ethical considerations. In 2023 5th international congress on human-computer interaction, optimization and robotic applications (HORA) . IEEE

2023
[91]

Monika Westphal, Michael Vössing, Gerhard Satzger, Galit B Yom-Tov, and Anat Rafaeli. 2023. Decision control and explanations in human-AI collaboration: Improving user perceptions and compliance. Computers in Human Behavior 144 (2023), 107714

2023
[92]

Mengting Xia. 2023. Co-working with AI is a double-sword in technostress? An integrative review of human-AI collaboration from a holistic process of technostress. In SHS Web of Conferences, Vol. 155. EDP Sciences, 03022

2023
[93]

It Might be Technically Impressive, But It’s Practically Useless to us

Qing Xiao, Xianzhe Fan, Felix Marvin Simon, Bingbing Zhang, and Motahhare Eslami. 2025. " It Might be Technically Impressive, But It’s Practically Useless to us": Motivations, Practices, Challenges, and Opportunities for Cross-Functional Collaboration around AI within the News...

2025
[94]

Vanshika Vats, Marzia Binta Nizam, Minghao Liu, Ziyuan Wang, Richard Ho, Mohnish Sai Prasad, Vincent Titterton, Sai Venkat Malreddy, Riya Aggarwal, Yanwen Xu, et al . 2024. A Survey on Human-AI Collaboration with Large Foundation Models. arXiv preprint arXiv:2403.04931 (2024)

2024 arXiv
[95]

Zhiguo Yang, Xiang Guo, and Peng Zhang. 2025. The New Normal and the Era of Misknowledge—Understanding Generative AI and Its Impacts on Knowledge Work. Knowledge 5, 4 (2025), 22

2025
[96]

Yu-Yin Wang and Yi-Shun Wang. 2022. Development and validation of an artificial intelligence anxiety scale: An initial application in predicting motivated learning behavior. Interactive Learning Environments 30, 4 (2022), 619–634

2022
[97]

Mamehgol Yousefi, Ahmad Shahi, Mos Sharifi, Alvaro J Jorge Romera, Simon Hoermann, and Thammathip Piumsomboon. 2025. Team Dynamics in Human- AI Collaboration: Effects on Confidence, Satisfaction, and Accountability. In Proceedings of the 27th International Conference on Multim...

2025
[98]

Man Zhang, Tao Yue, and Yihua He. 2026. Managing Uncertainty in LLM-based Multi-Agent System Operation. arXiv preprint arXiv:2602.23005 (2026)

2026
[99]

Yangqiaoyu Zhou, Haokun Liu, Tejes Srivastava, Hongyuan Mei, and Chenhao Tan. 2024. Hypothesis generation with large language models. In Proceedings of the 1st Workshop on NLP for Science (NLP4Science) . 117–139

2024
[100]

Zeda Xu, Chloe Soohwa Hong, Nicolás F Soria Zurita, Joshua T Gyory, Gary Stump, Hannah Nolte, Jonathan Cagan, and Christopher McComb. 2024. Adap- tation through communication: Assessing human–artificial intelligence part- nership for the design of complex engineering systems. ...

2024
[102]

Koji Yatani, Zefan Sramek, and Chi-Lan Yang. 2024. AI as extraherics: Fos- tering higher-order thinking skills in human-AI interaction. arXiv preprint arXiv:2409.09218 (2024)

2024 arXiv
[106]

Henry Peng Zou, Wei-Chieh Huang, Yaozu Wu, Yankai Chen, Chunyu Miao, Hoang Nguyen, Yue Zhou, Weizhi Zhang, Liancheng Fang, Langzhou He, et al
[107]

arXiv preprint arXiv:2505.00753 (2025)

Llm-based human-agent collaboration and interaction systems: A survey. arXiv preprint arXiv:2505.00753 (2025)

2025 arXiv
[2017]

Human factors 59, 2 (2017), 229–241

A closed-loop model of operator visual attention, situation awareness, and performance across automation mode transitions. Human factors 59, 2 (2017), 229–241

2017
[2023]

Business horizons 66, 1 (2023), 87–99

Artificial intelligence and knowledge management: A partnership between human and AI. Business horizons 66, 1 (2023), 87–99

2023
[2024]

arXiv preprint arXiv:2407.19098 (2024)

Evaluating human-ai collaboration: A review and methodological frame- work. arXiv preprint arXiv:2407.19098 (2024)

2024 arXiv
[2025]

European Journal of Information Systems 34, 2 (2025), 207–229

Design principles for artificial intelligence-augmented decision making: An action design research study. European Journal of Information Systems 34, 2 (2025), 207–229

2025

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.

[1] [22]

Robert Eccles. 2025. Hybrid Intelligence Teams: A Theoretical Framework for Human-AI Collaboration in Knowledge Work. A vailable at SSRN 5792345 (2025)

2025

[2] [1]

Habeeb Abdulrauf, Abdulmalik Adetola Lawal, Amarachi Nina Uma Mba, Com- fort Ademola, Zaynab B Yusuf, Shalewa Babatayo, and Idris Ayinde. 2025. Artifi- cial Intelligence in Journalism: A Narrative Review of Opportunities, Challenges, Ethical Tensions, and Human-Machine Collaboration. American Journal of Arts and Human Science 4, 4 (2025), 5–18

2025

[3] [2]

Manal Ahdadou, Abdellah Aajly, and Mohamed Tahrouch. 2024. Unlocking the potential of augmented intelligence: a discussion on its role in boardroom decision-making. International Journal of Disclosure and Governance 21, 3 (2024), 433–446

2024

[4] [3]

NIST AI. 2023. Artificial intelligence risk management framework (AI RMF 1.0). URL: https://nvlpubs. nist. gov/nistpubs/ai/nist. ai (2023), 100–1

2023

[5] [4]

Mohamed Adlan Ait Ameur, Erfu Yang, and Scott Brady. 2024. Advancing healthcare through intelligent human-robot collaboration: overview, challenges and opportunities. In 2024 29th International Conference on Automation and Computing (ICAC). IEEE, 1–6

2024

[6] [5]

Saara Ala-Luopa, Sami Koivunen, Thomas Olsson, and Kaisa Väänänen. 2024. Considerations on human-AI collaboration in knowledge work–recruitment experts’ needs and expectations. (2024)

2024

[7] [6]

Ying Bao, Xusen Cheng, Triparna De Vreede, and Gert-Jan De Vreede. 2021. Investigating the relationship between AI and trust in human-AI collaboration. (2021)

2021

[8] [7]

Giulia Bassi, Valeria Orso, Silvia Salcuni, and Luciano Gamberini. 2025. Un- derstanding Workers’ Well-Being and Cognitive Load in Human-Cobot Col- laboration: Systematic Review. Journal of Medical Internet Research 27 (2025), e75658

2025