Kerala High Court Becomes First To Ban AI In Judicial Decisions, Warns Judges Against Using ChatGPT

The policy makes it explicitly clear that judges are prohibited from using AI tools to deliver any form of legal judgment or order.

Popular AI tools such as ChatGPT and Deepseek are also off-limits unless specifically approved by either the Kerala High Court or the Supreme Court.

Source : Freepik

The Kerala High Court has become the first in India to introduce a comprehensive policy governing the use of Artificial Intelligence (AI) by members of the district judiciary and their staff. While the court acknowledges the growing presence of AI in administrative and research functions, it has drawn a firm line when it comes to core judicial responsibilities, reported Bar and Bench.

The policy makes it explicitly clear that judges are prohibited from using AI tools to deliver any form of legal judgment or order. “AI tools shall not be used to arrive at any findings, reliefs, order or judgment under any circumstances,” the High Court states, underscoring that the responsibility for judicial content rests solely with the judge.

Popular AI tools such as ChatGPT and Deepseek are also off-limits unless specifically approved by either the Kerala High Court or the Supreme Court. The guidelines emphasise that all cloud-based AI services should be avoided in order to protect sensitive information and uphold judicial confidentiality.

Ethical Guardrails and Human Oversight

To ensure ethical use, the policy mandates that any AI tool employed for official purposes must conform to key principles of transparency, fairness, accountability, and confidentiality. It permits the use of approved AI tools for routine functions like scheduling and court management—but with the caveat that such usage must be supervised by a human at all times.

Even when AI is used for translations or to generate legal references from databases, the results must be carefully reviewed either by judges themselves or qualified translators. The policy warns against overreliance on automation, noting that errors in citations or summaries can undermine judicial integrity.

“All legal citations or references, must be meticulously verified by the judicial officers,” it states, applying this caution even to court-approved tools that summarise case law or legal statutes.

Training, Monitoring, and Reporting Requirements

The High Court’s directive also places a strong emphasis on training. Judges and court staff must attend sessions organised by the Judicial Academy or the High Court to better understand the ethical, technical, and legal nuances of AI integration.

Each court is expected to maintain a detailed log of AI usage, including what tools were used and how their output was validated. “Courts shall maintain a detailed audit of all instances wherein AI tools are used,” the policy notes.

If a glitch or error is found in the output of any approved AI tool, it must be promptly reported to the Principal District Court. “The Principal District Judge shall forward the same to the IT Department of the High Court without delay so that appropriate safety reviews can be taken by the IT team immediately,” the policy instructs.

The policy is also applicable to interns and law clerks working under the district judiciary, ensuring uniform adherence across all levels of the system.

In effect, Kerala’s judiciary is signalling that while AI can be a useful assistant, it cannot, and must not, substitute the human conscience and legal acumen essential for delivering justice.