Effect of professional experience on automation bias

Investigate whether professional experience among practicing advocates and judicial officers naturally mitigates the automation bias observed among Indian law students when they evaluate and verify legal outputs generated by large language models, in order to inform long-term AI regulation for the Supreme Court and High Courts of India.

Background

The study’s second phase examines Indian undergraduate law students and reports substantial exposure to fabricated AI-generated citations, uneven verification practices, and limited formal training in the ethical use of AI. The findings suggest vulnerability to overreliance and automation bias, although the sample does not include practicing legal professionals or judicial officers.

The authors identify professional experience as an unresolved factor that may influence resistance to AI overconfidence and hallucinations. Addressing this question is presented as necessary for developing longer-term regulatory frameworks governing AI use in India’s higher courts.

References

Future research should expand this dual-audit methodology to practicing advocates and judicial officers. Investigating whether professional experience can naturally mitigate the “automation bias” observed in students will be essential for developing long-term regulatory frameworks for AI in the Supreme Court and High Courts of India.

— Can Legal AI Know When It Is Wrong? And Do Students Know When It Is?  (2608.21089 - John et al., 21 Aug 2026) in Section IX, “Future Work”