SBI Contemplates Developing Its Own Domain-Specific LLM

  • SBI plans to construct its own Large Language Model (LLM) to leverage its vast data reservoir.
  • Nitin Chugh, Deputy Managing Director, emphasizes the move towards domain-specific LLMs for enhanced independence.
  • The initiative underscores India’s pivotal role in AI development and the allure of data-driven endeavors.
  • SBI has proactively adopted robust governance models ahead of regulatory mandates.
  • Rajesh Nambiar warns of Generative AI’s transformative potential, particularly its impact on white-collar jobs.

Main AI News:

In a strategic move, the State Bank of India (SBI), the nation’s largest lender, is considering the development of its own Large Language Model (LLM) to harness the extensive data it possesses. Speaking at the Asia Economic Dialogue event, Nitin Chugh, Deputy Managing Director of SBI overseeing digital banking and transformation initiatives, revealed the institution’s intention to embark on this ambitious project.

Chugh emphasized the imperative for SBI to transition from reliance on open-source models towards fostering its proprietary domain-specific LLM. He articulated, “We should commence the development of our own large language models tailored to our specific domain. Relying indefinitely on open-source models might not suffice. We aim to either co-exist with open-source models or establish our own, subject to rigorous data sovereignty standards.

The envisioned LLM project holds the promise of independence for the bank, mitigating dependence on external technical frameworks. Chugh underscored India’s substantial contribution to the development of AI-based tools, with nearly half of engineers engaged in global corporations being of Indian origin. He stressed the allure of returning to India owing to the wealth of available data and highlighted the bank’s robust data infrastructure and talented pool of data scientists.

Prioritizing a proactive approach to governance, SBI adopted a robust model to oversee AI initiatives well ahead of anticipated regulatory frameworks such as the Digital Personal Data Protection (DPDP) Act, 2023, expected to be notified by mid-2024. Chugh underscored the regulator’s concerted efforts to address the evolving landscape, particularly in Generative AI, by aligning policies with market dynamics shaped by technological advancements.

At the same event, Rajesh Nambiar, Chairman and Managing Director of Cognizant India and chair of Nasscom, offered insights into the nuanced impact of Generative AI (GenAI). While acknowledging its short-term hype, Nambiar cautioned against underestimating its long-term influence, citing its transformative potential. He flagged concerns over its impact on white-collar jobs, particularly in Business Process Outsourcing (BPO) sectors, and highlighted the imperative for professionals to upskill in AI-enabled environments to remain competitive in the evolving job market landscape.


SBI’s venture into domain-specific LLM development marks a strategic shift towards self-reliance and data sovereignty. The bank’s proactive stance on governance and technological innovation positions it favorably in an evolving market landscape, while industry insights caution about the transformative impact of AI technologies on job markets, urging professionals to adapt and upskill to remain competitive.