During Q1 FY27 earnings calls, senior executives at Infosys, HCLTech, and Tata Consultancy Services (TCS) agreed that small language models (SLMs) are the technology moving the needle with enterprise clients, particularly in regulated sectors such as banking and financial services. This shift is driven by GPU supply constraints, lower operational costs, and hardware efficiency, as SLMs can run on edge devices. Gaurav Vasu, Founder and CEO of UnearthInsight, stated that no banks will adopt an LLM model due to data leakage risks, noting that enterprise-wide AI adoption globally remains under 10%. According to Marvel.ai, high-traffic enterprise deployments spend $50,000+ monthly on LLM APIs, whereas an equivalent SLM workload costs between $3,000 and $8,000. Infosys CEO Salil Parekh said SLMs are gaining traction in banking, built around the Finacle platform, where a smaller parameter set reduces running costs and improves token efficiency, with the same logic now extended to IT operations work.