The proliferation of Generative Artificial Intelligence has transformed benign cognitive offloading into a systemic risk of cognitive agency surrender. Driven by the commercial dogma of "zero-friction" design, highly fluent AI interfaces actively exploit human cognitive miserliness, prematurely satisfying the need for cognitive closure and inducing severe automation bias. To empirically quantify this epistemic erosion, we deployed a zero-shot semantic classification pipeline (τ=0.7) on 1,223 high-confidence AI-HCI papers from 2023 to early 2026. Our analysis reveals an escalating "agentic takeover": a brief 2025 surge in research defending human epistemic sovereignty (19.1%) was abruptly suppressed in early 2026 (13.1%) by an explosive shift toward optimizing autonomous machine agents (19.6%), while frictionless usability maintained a structural hegemony (67.3%). To dismantle this trap, we theorize "Scaffolded Cognitive Friction," repurposing Multi-Agent Systems (MAS) as explicit cognitive forcing functions (e.g., computational Devil's Advocates) to inject germane epistemic tension and disrupt heuristic execution. Furthermore, we outline a multimodal computational phenotyping agenda -- integrating gaze transition entropy, task-evoked pupillometry, fNIRS, and Hierarchical Drift Diffusion Modeling (HDDM) -- to mathematically decouple decision outcomes from cognitive effort. Ultimately, intentionally designed friction is not merely a psychological intervention, but a foundational technical prerequisite for enforcing global AI governance and preserving societal cognitive resilience.
Figures & tables
Figure 1: The Epistemic Skew and Agentic Takeover in High-Confidence HCI Literature (2023 – March 9, 2026). A deep-learning-based semantic stance classification of 1,223 highly confident, AI-related papers (filtered via a rigorous τ=0.7 threshold from an 8,000-paper ontological baseline retrieved via the OpenAlex API). The analysis reveals the persistent, systemic prioritization of the frictionless paradigm (optimizing efficiency and reducing cognitive load). Crucially, a granular semantic breakdown exposes a recent ”agentic takeover”: a brief 2025 surge in research defending strict human cognitive sovereignty (19.1%) was abruptly suppressed in early 2026 (13.1%) by an explosive shift toward optimizing machine autonomy and autonomous AI agents (19.6%). This dynamic structural divergence empirically quantifies the escalating marginalization of human epistemic agency in contemporary human-AI interaction design.
Friction Paradigm & Source
Cognitive Load Profile
Emotional & Metacognitive State
Epistemic Outcome
Extraneous Usability Friction (Cluttered interfaces, system latency)
Extraneous Load: Depletes limited working memory without serving the core epistemic task.
Frustration & Annoyance: Triggers negative affect and degrades sustained attention.
Task Abandonment: Comprehension failure and breakdown of human-AI collaboration.
The Zero-Friction Trap (Highly fluent, single-conclusion AI outputs)
Minimal Load: Requires almost no mental effort to process information.
Synergistic Synthesis & Sovereignty Defense: Forced awakening of System 2 analytical deliberation.
Table 1: Extended Taxonomy of Friction in Human-Computer Interaction. This taxonomy conceptually decouples scaffolded cognitive friction from traditional usability barriers and the zero-friction trap, mapping each paradigm to its distinct cognitive load profile, metacognitive state, and ultimate epistemic outcome.
Figure 2: The Two-Dimensional Evaluation Space Model for Cognitive Agency. The X-axis represents the delegation of sensemaking control (AI Dependency), while the Y-axis represents the depth of epistemic tension and analytical load (Cognitive Friction). Long-term interaction within the Cognitive Agency Surrender quadrant drives systemic cognitive atrophy. In contrast, Synergistic Synthesis utilizes structured friction to break the zero-sum fallacy of automation, preserving epistemic sovereignty through Meaningful Human Control (MHC).
Figure 3: Causal Pathway of Human-AI Synergistic Synthesis. Unlike conventional consensus-driven workflows (upper section) that promote premature cognitive closure and heuristic acceptance, the scaffolded cognitive friction framework (lower section) utilizes heterogeneous agents and a Devil’s Advocate to expose structured contradictions. This injected epistemic tension elevates visual sampling entropy, forcibly awakening System 2 analytical reasoning to adjudicate logical trade-offs and preserving the human decision-making locus.
Appendix figures & tables2 assets
Supplementary material from the paper’s appendix.
Appendix
Analytical Tier
Deep Learning Semantic Label (Stance Description)
1. The Frictionless Paradigm (Efficiency & Usability)
”optimizing system efficiency, seamless interaction, and reducing user effort”
2. Machine Autonomy (AI Agents & Reflection)
”improving machine autonomy, artificial intelligence agents, and algorithmic reflection”
3. Strict Human Agency (Epistemic Sovereignty)
”preserving human cognitive agency, critical thinking, and preventing automation bias”
Appendix
Table A1: Semantic Labels for Zero-Shot Stance Classification. The exact descriptive strings fed into the BART-large-MNLI model to compute the contextual ideological distribution within the AI-HCI corpus.
Figure A1: Normalized Confusion Matrix of Zero-Shot AI vs. Expert Goldline. Colors represent class-wise recall (accuracy per category), effectively neutralizing visual biases caused by extreme class imbalances. The pipeline demonstrates exceptional sensitivity (84.6%) for the marginalized Strict Human Agency paradigm. Overall pipeline accuracy is 87.0% evaluated on the 200-paper stratified sample.
AI agents pose significant risks as they are granted increasing autonomy. A commonly proposed solution is human oversight and keeping a ''human in the loop'', but this is not a simple solution: Not only do current approaches to AI agent design impede effective human oversight, but the cognitive capacities required for it are also themselves degraded by extended use of AI systems. This position paper argues that current approaches to the development and deployment of AI agent systems do not support effective human oversight -- they contribute to its degradation. To address this, a top priority in the advancement of AI agents should be supporting the situated goals and cognitive requirements of effective human oversight, treating the human needs of overseers at the same level of importance as AI agent capability. To put this idea into practice, we connect work on automation and human-computer interaction to AI agent processes, outlining design-level affordances and organizational protocols that (1) support overseers in exercising critical judgement and (2) counteract the skill atrophy that arises from extended use of automation. We urge developers and deployers to adopt these or similar approaches. Without explicit support for the cognitive demands of effective human-agent interaction, AI agent systems will continue to passively incentivize the degradation of the very human skills they rely on.
This paper examines three recent frameworks for understanding the cognitive and epistemic consequences of artificial intelligence: Tri-System Theory, Thinkframes, and System 0. It argues that while the first two capture important dimensions of AI's influence on individual reasoning and collective epistemic practices, System 0 occupies a theoretically distinctive position that neither can fully replicate. The paper introduces the concept of cognitive colonization, according to which AI systems can embed external interests within the architecture of the self in ways that are difficult for users to perceive. Because such systems are already widely deployed, understanding these invisible forms of influence is an urgent philosophical and practical task.
Frontier agentic systems powered by large language models (LLMs) exhibit human-like patterns of cognition. As these systems become deeply integrated across different domains, their cognitive engagement raises critical concerns for human society that remain insufficiently studied. To address this gap, we systematically analyze risks induced by expanding cognitive capabilities, following a three-level framework defined by their cognitive scope, from physical cognition to social cognition, and finally to self-referential cognition. We study their potential risks to human agency, autonomy, and control capability, corresponding to each cognitive level. We finally propose strategies to mitigate these risks and enhance the controllability of agentic AI systems, ensuring their long-term safe development.
Guanchu Wang, Qinuo Li, Mengnan Du +2
Shanghai Artificial Intelligence Laboratory, Shanghai, China · The Chinese University of Hong Kong, Shenzhen, China