Sebi developing LLM to cut approval times further: Narayan

Mumbai: Capital markets regulator Sebi is training a large language model (LLM) to further cut processing times on approval requirements, whole time member Ananth Narayan G said on December 18.

“We are actually starting to train an in-house LLM in Sebi to process, so that the processing times for fresh funds and fresh schemes can come down dramatically,” Narayan said, addressing a CII event here.

LLMs are advanced AI systems designed to understand and generate human language. They use deep learning techniques and are trained on massive amounts of text data, allowing them to perform tasks like writing, summarizing, and translating text.

While efforts to deploy AI and machine learning are on in all sectors, this may be the first time that a financial regulator in the country would have disclosed specific efforts like developing an LLM.

Narayan said Sebi has already reduced the time taken for processing applications for clearance on mutual fund schemes and initial public offerings to less than three months, and aims to reduce it further to under a month.

The trained LLM will make it “easier” and “scientific” to achieve the target of reducing the processing times, he said.

“The training is already underway and the effect should show up soon,” he said.