Now Reading: Sebi plans AI surveillance of corporate filings to catch fraud
-
01
Sebi plans AI surveillance of corporate filings to catch fraud
Sebi is expanding its AI-based surveillance from trading data to corporate filings to proactively detect financial misstatements and manipulation in quarterly results, whole-time member Kamlesh Chandra Varshney said at a Ficci conference. The regulator has built a dedicated team and an AI model for this, moving beyond the current complaint-driven corporate investigations.
Existing tools include R(AI)DAR for reviewing misleading ads, Project Sudarsan for real-time social media fraud detection (which has flagged over 20,000 instances), and the InfoMerge system that automates investigation tasks and cuts processing time. Sebi also plans to extend its AI ad review beyond asset management companies and expand outreach with Meta and Google to flag fraudulent content.
Separately, a Tofler report says AI is making corporate due diligence faster for Indian businesses by aggregating data from over 100 sources covering 10 million firms. Tools like ToflerGPT enable conversational search, and the Tofler Score uses a 62-point framework to assess business risk, particularly helping small and medium enterprises that lack dedicated compliance teams.
Indian Opinion Analysis
Rediff.com led with Sebi's specific enforcement expansion and named the official and the conference, framing it as a proactive regulatory upgrade. Times Now led with a general industry report on AI due diligence, omitting Sebi's new tool entirely. The two stories cover different subjects, so no stance conflict arises. The balanced reading is that both point to AI's growing role in corporate oversight, but Sebi's move targets listed company filings specifically, while Tofler's tool serves broader business verification. The key watch item is Sebi's implementation timeline for its corporate filing AI model.
Coverage: 2 sources, 2 neutral
Sources (2): rediff.com (neutral report), timesnownews.com (neutral report)
This brief was synthesised by AI from the 2 sources linked above, so one read covers every framing they carry. Methodology and corrections.
Updated: this story now draws on 2 sources.
