On June 3 the Supreme Court released a small preliminary report - of 35 pages only - that never quite made the headlines. It should have. The release of the Draft Regulations for Use of Artificial Intelligence in Courts was a landmark moment - an acknowledgement of the relentless intrusion of AI and a proposal on working with, rather than against, the technology.
The proposed regulations - offered to the public for study, review, and recommendations till July 15 - urge courtrooms to "actively seek opportunities" to deploy AI-based systems and tools to improve access to justice, or in some cases ensure it, and enhance operational efficiency by using it to automate administrative functions, like transcription and translation.
Most importantly, they also offer safeguards against misuse of AI, including it becoming the final arbiters of justice.
The draft rules can be accessed here and recommendations submitted to the court on this email - office.regcc@sci.nic.in.
The goal, the report said, is to "establish an institutional framework for responsible AI adoption" while ensuring its use remains "grounded in the principles of human primacy, transparency, accountability, data protection, and judicial independence".
This architecture is meant to ensure AI remains "strictly subservient to human judgment and judicial authority".
Equally important, the regulations allow for phased integration, which includes delinking adoption by the top court and high courts. That means each court can adopt the rules, or parts of them, on different dates, allowing for integration based on what is needed and possible rather than a forced one-size-fits-all model.
The 35-page report also outlined the "permissible" and "prohibited" uses.
In the latter, arguably more important, category were restrictions on the use of personal data to train AI models and the use of those models in verdicts, including bail hearings.
Red flags were also raised over the use of AI as "an independent source of evidence without full and transparent disclosure of its character". This is particularly important in cases where the prosecution or defence might present AI-generated images or videos as illustrative examples.
The report also proposed an oversight mechanism - an AI Committee - to be set up at the Supreme Court and High Court levels, along with a "permanent, full-time body" for policymaking and institutional oversight.
The latter, the report said, shall consist of Supreme Court judges and two High Court Chief Justices nominated by the Chief Justice of India, in addition to a high-ranking civil servant from the Ministry of Electronics and Information Technology, domain experts in AI, finance and cybersecurity, and senior legal professionals.
This full-time body will consist of five subcommittees - Judicial, Technical, Infrastructure and Finance, Case and Data Management, and Cybersecurity.
Finally, in the interest of transparency, all litigants will be informed of the use of AI in their cases, though that information will be restricted to instances of "material assistance" only.
The Supreme Court last month, meanwhile, expressed grave concern over a National Company Law Tribunal (NCLT) judgment that relied on fake precedents and cases generated using AI. While setting aside the verdict, the court observed that human control must remain absolute at every stage and warned against AI making legal professionals dependent.
RECAP | "Invisible, Insidious": Top Court Warns Against Use Of AI In Legal Process
"This byproduct, that is AI, that is the production of fake, non-existing, and hallucinated material, and its utilisation, as precedents in law, is like release of methyl isocyanide within the province of law and justice. Invisible, insidious, and catastrophic, by the time anyone notices, it is not only contaminant, but takes away the very lifeline of judicial determination," it added.
