70% of Indian Banks Run AI in Production as Security Concerns Limit Scale: Zeta

Indian financial institutions have successfully moved artificial intelligence from experimentation into production, but scaling these technologies remains a significant hurdle. A new survey of 40 CXOs across 18 leading banks and NBFCs reveals that while 70% have deployed AI selectively, the transition to repeatable, institution-wide implementation is currently constrained by infrastructure and control gaps.

What was announced

Zeta released the findings of its 2026 CXO survey, which highlights a growing divide between successful AI piloting and repeatable, scaled deployment within the Indian banking sector. While 30% of Chief Data Officers (CDOs) report that their institutions have reached scaled deployment, the majority of adoption remains concentrated in structured, reviewable workflows. Retail lending leads the way, with 88% of Chief Operating Officers (COOs) identifying it as an area of meaningful impact, followed by customer service at 75% and back-office operations at 63%.

The technical barriers to scaling are shifting from basic connectivity to sophisticated usability. Although 80% of CIOs and CTOs describe their data environments as "mostly ready" for AI, none consider them "fully ready." Key constraints include insufficient training data (61%), privacy and consent issues (53%), and data silos (46%). To combat this, 67% of banks are currently using or piloting AI to enrich their own data sets. Furthermore, while 79% of institutions have adopted real-time data platforms and API-first architectures, only 43% have implemented the advanced analytics and MLOps capabilities required to manage AI workloads consistently.

The survey also notes that AI has a strong foothold in software engineering, with 80% of technology leaders using it for testing and QA. However, adoption thins as the technology moves from production to execution; only 30% use AI for deployment or incident detection. Security and data privacy remain the primary barriers to deeper integration, scoring 3.89 out of 5, whereas a lack of ROI clarity was rated as the lowest barrier at 2.0 out of 5.

"Indian banks have shown that AI creates value in production. The next challenge is making that success repeatable, and the survey is clear about what stands in the way: not conviction, but control," said Sivaram Kowta, President, Zeta India. "Banks have connected their core systems. The next step is to make them usable by AI, with banking context and permissions built in, and to put in place the identity, policy and audit controls that let risk and security leaders say yes with confidence. Banks that build these foundations once will find that the tenth deployment costs a fraction of the first."

Sivaram Kowta, President, Zeta India.

The companies involved

Zeta is a provider of next-generation banking technology, positioning itself as a modern alternative to legacy core systems. The company operates globally, including a significant presence in India through Zeta India, where it works with major financial institutions to modernise their technology stacks. Zeta’s platform is designed to be API-first and cloud-native, catering to banks and non-banking financial companies (NBFCs) looking to transition away from monolithic architectures. In the current market, Zeta competes by offering modular banking suites that allow for faster product launches and more granular control over data and customer experiences.

The firm has established itself as a key player in the "banking-in-a-box" and core-as-a-service sectors, focusing on the intersection of cloud computing, real-time data processing, and the integration of artificial intelligence into the banking core. Its market position is defined by helping incumbent lenders face pressure from fintech challengers by providing the infrastructure necessary for digital-first banking services. Zeta’s focus on "banking context" within its data layers is a specific differentiator in a market where many institutions struggle with the usability of their existing data silos.

What FF News has reported before

FF News has followed Zeta’s rapid ascent in the fintech space, notably reporting on the company’s financial milestones and product expansions. In early 2025, the publication covered how the firm was Zeta Valued at $2 Billion in New $50 Million Strategic Fundraise, a move that solidified its unicorn status and provided the capital necessary for further technological development. More recently, FF News detailed the company’s direct entry into the generative AI space with the report Zeta Launches Selene: Gen AI Customer Support Agent for Banks and Fintechs. This product launch highlighted Zeta’s strategy of moving beyond core infrastructure into specific AI-driven applications designed to automate complex customer interactions and support workflows for financial institutions.

What this means

The findings suggest that the "honeymoon phase" of AI experimentation in banking is ending, replaced by the sober reality of architectural debt. While banks have proven AI’s utility in isolated pockets like retail lending and QA testing, the lack of a unified "AI-ready" core is creating a bottleneck. The industry is currently facing a choice: continue building expensive, siloed AI solutions or invest in a fundamental overhaul of data permissions and governance layers. As institutions move toward consequential decision-making in credit and fraud, the pressure on legacy core systems will intensify. The real winners in this next phase will not necessarily be the banks with the most advanced models, but those that first solve the "repeatability" problem through standardised engineering controls.