Reconceptualizing Community of Inquiry in the Age of Generative Artificial Intelligence

Synthesis: Ba, Gašević, Lim, and Anderson (2026) reconceptualize the Community of Inquiry (CoI) framework for the age of generative AI, arguing that GenAI challenges the assumption that indicators of presence can be attributed primarily to human learners and instructors. Rather than framing GenAI as a mere tool, dialogic partner, or 'fourth presence,' they position GenAI as an epistemic condition that shapes inquiry through direct use, indirect mediation of tasks, and latent influence on how contributions are interpreted and evaluated. Drawing on sociomaterial and postdigital perspectives, they argue that CoI presences are sociotechnical accomplishments emerging through human-GenAI assemblages, and propose a configuration-based heuristic where the relationship between GenAI involvement and inquiry quality is conditional on human accountability.

Key Findings

GenAI can produce fluent discourse resembling strong cognitive, social, or teaching presence without reflecting learners' actual thinking, relationships, or regulatory intentions.
The paper reinterprets cognitive, social, and teaching presence as sociotechnical accomplishments of human-GenAI assemblages rather than purely human activity.
It proposes a configuration-based heuristic suggesting the relationship between GenAI involvement and inquiry quality is conditional on human accountability: strong presence can occur with high or low GenAI involvement when accountability is strong, and weak presence with either when accountability is weak.
CoI evaluation needs process-sensitive evidence (prompting and revision traces, disclosure and attribution practices, verification moves, interaction logs) rather than inference from polished final texts.

The CoI Framework Under GenAI Pressure

CoI conceptualises a meaningful educational experience as emerging from the dynamic interplay of cognitive, social, and teaching presence, grounded in social Constructivist and Deweyan pragmatic traditions. GenAI destabilises these presences in two ways. It can inflate participation that linguistically resembles CoI indicators while weakening the underlying inquiry processes, and it changes the scale, fluency, opacity, and detectability of performative participation relative to older concerns such as lurking or superficial engagement. While prior adaptations of CoI added presences (e.g., learning, emotional, autonomy presence), this paper argues for a more substantive reconceptualization: the framework's constructs and operational indicators need reconsideration under conditions where machine-produced discourse can masquerade as authentic inquiry.

Why GenAI Is an Epistemic Condition

The paper critiques three prior framings of GenAI in CoI — as a tool/infrastructure, as a participant/dialogic agent, and as a speculative 'fourth presence.' Each is useful but insufficient: the tool framing under-theorises systemic epistemic shifts; the dialogic-agent framing amplifies explicit participation but misses indirect and latent influence; the fourth-presence framing risks reifying GenAI as a co-equal agent despite its lack of intentionality and epistemic responsibility. Instead, the paper positions GenAI as an epistemic condition — a pervasive influence that reconfigures how presences are enacted, interpreted, evidenced, and governed. It operates through a visible dimension (the outputs learners engage with, which embed stylistic and epistemic norms) and an invisible dimension (training data, algorithms, platform logics, embedded biases). The pedagogical decision thereby shifts from whether GenAI is available to how the inquiry environment is structured so that attribution, verification, and responsibility remain visible.

Reinterpreting the Three Presences

Cognitive presence now unfolds where GenAI increases the availability of fluent explanations, accelerating early sense-making but heightening the risk of premature closure when coherence is mistaken for warrant. Each phase of the practical inquiry model (triggering event, exploration, integration, resolution) is enacted through distinctive human-GenAI assemblages: machine-generated suggestions can catalyse triggering, exploration expands rapidly but requires interrogating algorithmic origins, integration risks machine-drafted syntheses that obscure inaccuracies or pull groups toward homogeneous formulations, and resolution becomes a test of whether claims are accepted because compelling or because warranted.

Social presence is reshaped as machine-produced utterances mimic friendliness, empathy, and responsiveness with high linguistic credibility, complicating relational accountability. Affective expression, open communication, and group cohesion remain relevant but their enactment changes: supportive messages may be less credible if perceived as delegated, automated active listening may weaken the sense of genuine attendance, and stylistic convergence may reduce the visibility of individual voices. Explicit labelling of AI-assisted contributions can preserve social presence by reducing authorship uncertainty.

Teaching presence shifts from delivering content to sustaining regulative functions that keep inquiry coherent, equitable, and epistemically disciplined. Design and organization now include deciding how GenAI is positioned in tasks; facilitating discourse includes modeling the questioning of generated outputs and preventing AI-smoothing from erasing productive conflict; and direct instruction becomes ongoing sense-making infrastructure for evaluating machine-generated information, detecting hallucinated citations, and recognising bias. Teaching presence becomes a distributed accomplishment rather than an individual instructor capacity.

Conditional Configurations and Practical Design

A two-axis heuristic (degree of GenAI involvement; strength of human accountability) yields four quadrants illustrating that the relationship between GenAI involvement and CoI presence is conditional, not linear. Pedagogically, the paper argues against assuming learners already know how to engage in inquiry under GenAI conditions, proposing Simulation-based training with scripted roles and GenAI decision points as a teachable progression toward independent inquiry. Distributed teaching presence can be enacted through structured checkpoints, rotating roles (evidence-checker, counterexample generator, bias-spotter), and warrant-check routines (what claim is made, what its basis is, what is missing, what would confirm it). Assessment should shift from polished outputs to evaluating process and growth in capability, using prompt-response logs, revision histories, structured rationale notes, attribution statements, and verification moves — while respecting that instructors often lack the access or ethical justification to collect private prompt logs, favouring learner-submitted artifacts and institutionally governed tools.

Implications for Theory and Research

CoI remains theoretically valuable but its operationalizations need refinement: existing coding schemes and surveys risk over-attributing presence based on surface features of text. The paper sets a research agenda of refining CoI instruments for GenAI-mediated environments, testing the configuration-based hypothesis across real human-GenAI assemblages, and examining simulation-based training mechanisms. It is a conceptual analysis rather than an empirical study, so its claims are theoretically grounded propositions for future investigation, developed mainly with higher-education, text-based inquiry contexts in mind.

Connected Concepts

Community Of Inquiry — Community of Inquiry

Connected Articles

AI Communities Of Inquiry 2026 — AI in communities of inquiry
AI Online Education Engagement Satisfaction 2026 — AI and online education engagement
Critical Thinking GenAI Scaffolding — Scaffolding critical thinking with GenAI

Citation

Ba, S., Gašević, D., Lim, C.-P., & Anderson, T. (2026). Reconceptualizing community of inquiry in the age of generative artificial intelligence. Computers and Education: Artificial Intelligence.

View Source View Changes Edit