Kerala High Court bars AI in judgments, issues strict guidelines for district judiciary

The Kerala HC’s policy follows Chief Justice B.R. Gavai’s warning against unchecked generative AI use, mandating that any AI-related errors be reported to the Principal District Judge and forwarded to the High Court’s IT Department for review

In a significant step towards regulating artificial intelligence (AI) within the legal system, the Kerala High Court has released a policy framework outlining clear boundaries for AI use across the state’s district judiciary. The new guidelines stress caution, responsibility, and ethical integrity in the application of AI tools for judicial tasks.

The policy covers a wide range of AI technologies, including generative models and AI-driven legal databases used for accessing case law, statutes, and other resources. While acknowledging the growing presence of AI in legal processes, the High Court has strictly prohibited its use in making judicial determinations. Judges are barred from relying on AI tools to deliver findings, grant relief, or issue judgments under any circumstances.

“The final responsibility for judicial decisions lies solely with the presiding judge,” the policy notes, highlighting concerns around accountability and the risk of AI-generated inaccuracies. Courts are now required to maintain a detailed audit trail of all AI usage, documenting verification steps taken by human users to check the accuracy of results.

Confidentiality and accuracy prioritized

The guidelines emphasize that AI usage must uphold judicial principles such as fairness, transparency, and confidentiality. Judges and court staff are expected to verify all AI-generated content — including citations and legal references — before using it in any capacity. This directive applies to all approved tools, including databases enhanced by AI technologies.

Notably, the use of public generative AI platforms like ChatGPT and DeepSeek is explicitly banned. The High Court cautioned that uploading sensitive case data to cloud-based tools could compromise confidentiality and risk misuse by service providers to train or improve their models.

To foster informed and ethical use, the Court also directed all judiciary members and supporting staff to attend training sessions organized by the Judicial Academy or the High Court. These will address the legal, practical, and ethical implications of AI in judicial work.

The move follows earlier remarks by Chief Justice of India B.R. Gavai, who in March warned against the unchecked use of generative AI, citing incidents of fabricated citations and legal inaccuracies. Kerala’s new policy also mandates that any AI-related errors be reported to the Principal District Judge and escalated to the High Court’s IT Department.

The Court reaffirmed that empathy, ethics, and context — essential components of justice — remain beyond the scope of AI.