evidoria

← Back to browse

Good practice Imported

Project SUDARSAN & R(AI)DAR — India's SEBI Uses AI to Flag Over 100,000 Pieces of Misleading Financial Content

India · Mumbai · See the India profile · See the Mumbai profile

Evidence: Descriptive / self-reported Top 62% 53/100 · Ask Evidence Copilot about this practice

SEBI's AI tools SUDARSAN and R(AI)DAR scan social media and ads for financial fraud: Chairman Tuhin Kanta Pandey says SUDARSAN has identified 20,000+ fraudulent posts in real time and SEBI has flagged 100,000+ misleading items to platforms, against 550 crore daily trade messages.

20,000+
Fraudulent posts identified by SUDARSAN (to Aug 2026)
100,000+
Misleading content items flagged to platforms (to Aug 2026)
~550 crore
Daily order/trade messages (context)
Project SUDARSAN & R(AI)DAR — India's SEBI Uses AI to Flag Over 100,000 Pieces of Misleading Financial Content

Details

Maturity
Established
Promoter
Securities and Exchange Board of India (SEBI)
Period
2025–2026
Keywords
financial regulation, investor protection, market surveillance, fraud detection

Context

India's Securities and Exchange Board of India (SEBI) has moved from case-by-case enforcement to always-on AI screening of financial content, via Project SUDARSAN (scanning videos/social media for fraud markers) and R(AI)DAR (reviewing advertisements for misleading claims).

Activities

SUDARSAN and R(AI)DAR screen social media, video and advertising content at scale, against a backdrop of roughly 550 crore daily order and trade messages. SEBI is also drafting a formal AI/ML governance framework mandating human oversight, kill-switch controls and tiered accountability.

Results

SEBI chairman Tuhin Kanta Pandey said SUDARSAN identified more than 20,000 instances of fraudulent content in real time, and that SEBI has flagged over 100,000 pieces of potentially misleading content to social media platforms.

Conclusions

SEBI has run comparable social-media-monitoring efforts before with mixed follow-through; no public precision, false-positive or enforcement-conversion rate has yet been published for either tool.

Implementation
Indicative cost
High (€500k–€5M) — Not disclosed.
Time to results
Long (> 3 years) — Ongoing, always-on screening; governance framework still being drafted as of August 2026.
Staffing & skills
SEBI surveillance and enforcement teams

Conditions for success

  • Continuous, always-on screening rather than periodic case-by-case review
  • A formal AI/ML governance framework with human oversight and kill-switch controls

Common failure modes

  • No independently audited precision or false-positive rate published
  • SEBI has run comparable monitoring efforts before with mixed follow-through

Commonly funded by

National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful