How SEBI Is Using AI To Track Suspicious Trades In Real Time

India’s market regulator is expanding its use of artificial intelligence to monitor trading activity and identify unusual patterns that could point to market manipulation. Its systems are being used to analyse trading and social media data and flag activity that may require further investigation.

According to a Bloomberg report, Securities and Exchange Board of India (SEBI) analysts are using AI and machine-learning tools to examine stock volumes, derivative positions and algorithmic trading identifiers for unusual activity. The regulator is also tracking stock tips and promotional posts on social media and forums and comparing them with trading activity to identify possible pump-and-dump schemes.

The technology was also used in a recent enforcement action involving a JPMorgan Chase unit. Surveillance systems reportedly flagged suspicious activity linked to a newly introduced closing auction mechanism, with SEBI taking action six days after the activity was detected. An alert generated through AI and machine-learning systems was among the factors considered in the case.

SEBI Expands AI-Based Market Surveillance

SEBI’s use of AI is part of a broader effort to strengthen automated surveillance as trading strategies become more complex. Machine learning has helped the regulator identify unusual trading patterns more efficiently and has reportedly reduced the time taken for some parts of the initial public offering review process by as much as 70 per cent.

The regulator has also expanded its technology and surveillance workforce. Its team of engineers and analysts has grown to more than 200 employees from around 20 five years ago.

SEBI began investing in its own data centre in Mumbai in 2019, with the facility becoming operational in 2021. That was also the year the regulator started deploying its in-house AI system, Sudarshan.

The early versions of the technology reportedly faced problems with large language models, including inaccurate outputs. SEBI also faces challenges in ensuring that the data used by its AI systems is reliable.

AI Also Tracks Social Media Activity

SEBI has expanded its use of technology beyond market data to social media, where misleading investment-related content can reach a large number of retail investors.

The regulator makes as many as 7,000 requests every month to platforms including X, Instagram and Telegram to remove misinformation, Bloomberg reported. The system has also helped remove more than 100,000 videos across different networks.

While social media data can also create challenges for AI systems, particularly when content is generated by bots or contains manipulated information.

Despite the wider use of AI, the technology does not independently decide whether regulatory action should be taken. Officials continue to review AI-generated outputs and make adjustments before decisions are taken.