The Supreme Court of India came down heavily against the use of artificial intelligence in court filings on Tuesday — calling it a wholly unnecessary practice. The bench headed by Chief Justice Surya Kant also voiced serious concern about the growing trend of lawyers filing petitions drafted with AI tools that contain non-existent judgements, hallucinated quotes and biased information.
“We are alarmed to reflect that some lawyers have started using AI to draft petitions. It is absolutely uncalled for,” the bench said.
The remarks came mere days after the apex court dismissed a Special Leave Petition and orally cautioned all lawyers to exercise due diligence. The bench comprising Justice BV Nagarathna and Justice Ujjal Bhuyan also stressed the need for the bar to verify whether the details included in petitions were real or a deepfake.
Growing trend of ‘fake’ cases
The Supreme Court had encountered its first major case of AI misuse in December last year — after a petitioner used it to draft a response. According to a Times of India report, the document had contained ‘hundreds of fake cases’ and left the lawyers red faced upon detection. Lawyers appearing for Omkara Asset Reconstruction had further claimed that fake cases were cited to convince the court in favour of Gstaad Hotels.
The issue appears to have gained momentum in the ensuing weeks — with Justice Nagarathna telling the apex court on Tuesday that she had recently come across a non-existent citation for ‘Mercy vs Mankind’. She added that in some cases the judgements were broadly accurate with AI attributing fake quotes within those verdicts and making it extremely difficult to verify the contents.
Chief Justice Surya Kant also referred to a similar instance — noting that Justice Dipankar Datta’s court had recently seen “not one but a series of such judgements” being cited.
Does the Supreme Court use AI?
Interestingly enough the Supreme Court itself has adopted limited use of AI in its daily functioning — now using technology for administrative and supportive functions. CJI Kant had reconstituted an AI Committee in December 2025 to oversee such functions under strict caution to avoid biases or errors. The four-member committee headed by Justice Narasimha looks at ‘initiatives relating to the adoption, development and deployment of AI tools in the judiciary’.
AI-generated lawyers, legal LLMs and tech startups
Matters have taken a dystopian turn in courthouses across the world — with many now turning to AI avatars and digital ‘lawyers’. Large language models have also ventured into legal systems and there is a growing number of AI-driven startups working on cases. These companies and tools currently focus on case analysis and filings — often analysing and preparing the materials that will be used in court by human lawyers.
A case from March 2025 saw Jerome Dewald using an AI-generated avatar to represent himself in an employment dispute while appearing remotely before a New York court. The reaction has been somewhat mixed — with some US courts embracing artificial intelligence in the legal space while others remain vocal critics of the practice. Months later a Phoenix family court allowed an AI video of a deceased victim while delivering a sentencing message in May 2025 — potentially swaying the judge toward a maximum sentence.