Indian legal practice has huge AI demand — driven by court backlog — but also a sharp, specific risk: AI hallucinating case law. Here’s a grounded view. (dgm implements osFoundry, a separate company’s platform — dgm is an independent integration partner, not osFoundry, and this is not legal advice.)
The use cases
- Contract review and due diligence — reviewing thousands of pages quickly.
- Legal research — conversational research and precedent discovery via semantic search.
- Drafting — contracts and documents.
- Litigation support — and court-side tools like Adalat AI for transcription across thousands of courts.
Large firms (Shardul Amarchand, Cyril Amarchand) have deployed or piloted firm-wide GenAI, and judicial pilots like the Supreme Court’s SUPACE are emerging.
Backlog drives demand
India has an enormous court backlog — cited in the tens of millions of pending cases — creating strong demand for efficiency. Senior judiciary figures have suggested AI could help resolve a significant share of it, and government and court pilots are exploring AI assistance. The opportunity is real.
The defining risk: hallucinated citations
Here’s the legal-specific danger: citing AI-hallucinated, non-existent judgments. The Supreme Court flagged this as “alarming” in early 2026, and using fabricated AI citations can amount to professional misconduct. So every AI-assisted legal output must be verified against authoritative sources before use — AI accelerates research, but the lawyer remains responsible for accuracy. This isn’t optional; it’s the line between useful AI and a sanction.
A regulatory vacuum — so self-govern
As of 2026, the Bar Council of India has no formal AI guidelines — a regulatory vacuum with no specific disclosure or data-residency obligations. So firms must self-govern client confidentiality, accuracy and verification, applying their own discipline until formal rules arrive (a committee is expected following the Supreme Court’s attention).
Where osFoundry fits
osFoundry orchestrates legal AI — research, contract review, drafting — with retrieval grounded in authoritative sources (supporting verification), self-hostable for client confidentiality. The discipline of citing sources and verifying is built into how dgm designs it. dgm builds the controls; lawyers own professional responsibility. osFoundry is younger with limited independent coverage, so dgm validates fit.
How dgm helps
dgm builds legal AI on osFoundry — contract review, research, drafting — with the client-confidentiality and citation-verification discipline legal work demands, self-hosted where needed. Transparent pricing: $399 assessment, $3,999/month implementation, no per-seat fees (INR approximate; 18% GST for domestic clients). Explore the platform at osFoundry, or talk to dgm about legal AI.
General information, not legal advice. AI outputs must be verified; lawyers remain responsible for accuracy and confidentiality.