Author
Assoc Prof Suvrajyoti Gupta
Organisation/Institution
OP Jindal Global University
Country
INDIA
Panel
Legal Education
Title
“IN A MACHINE DARKLY”: AI AS A TOOL FOR REFLECTIVE PEDAGOGY IN LAW SCHOOL
Abstract
This paper argues that artificial intelligence can be reframed from a mere informational instrument into a rigorous, ethically governed reflective partner for professional formation. Drawing on Donald Schön’s concept of reflective professionalism (1983) and its influence on professional identity, the paper acknowledges both its pedagogical value and its major criticisms. Eraut (1994) contends that Schön’s reflection-in-action romanticizes practice by ignoring the tacit, intuitive knowledge underlying expertise. Newman and Moon (1999) highlight the model’s conceptual ambiguity, noting that terms such as knowing-in-action and reframing lack operational precision. Boud and Walker (1998) and Brookfield (2017) further argue that reflection is shaped by power, hierarchy, and affect, yet Schön treats it as an isolated cognitive act detached from social context. Finally, reflective classrooms are vulnerable to ‘groupthink’, which Janis (1972) defines as a state where “members’ striving for unanimity overrides their motivation to realistically appraise alternatives. “To address these weaknesses, the paper proposes a reflective-AI pedagogy built on machine-learning and large-language-model systems in three calibrated forms: (a) Aligned AI, a neutral, bias-tested mirror that helps identify cognitive barriers and conformity; (b) AI Avatars, modelled on each student’s cognitive, emotional, and linguistic profile, functioning as external Socratic mirrors that convert internal reflection into structured dialogue; and (c) misaligned mildly AI, systems with limited, controllable divergence from social or ethical norms, used to “crash-test” assumptions through counter-intuitive or norm-violating propositions that deepen epistemic resilience. The deployment of such tools requires strict ethical guardrails: multilayered content filters, provenance logging, human-in-the-loop moderation, curricular gating, anonymisation, and opt-out options. These would be overseen by an aligned guardian system— “Chuck Norris”—empowered to suspend interactions that drift from genuine reflection into manipulation or conspiracy. AI, thus reconceptualised, becomes not a threat to judgment but a calibrated mirror and stress-test for professional learning in a technology-saturated age. Though the paper is relevant to all professional education it confines itself to the instance of legal professional education.
Biography
Prof. Suvrajyoti Gupta is an Associate Professor at JGLS. He completed his first degree in law from the NUJS, Kolkata in May 2006. Between 2006-2009, he practiced as an appellate litigator, primarily based in the Supreme Court of India and the High Court of Delhi. His practice areas included commercial litigation, arbitration, constitutional and criminal law. Between 2009-2010 Mr. Gupta has successfully completed his LL.M. in International Business Law from the National University of Singapore. He has been a professor of law in OP Jindal Global University with several years standing, his research interests include , dispute resolution, AI and legal education.
