State Bank of India (SBI) is preparing to significantly expand the use of artificial intelligence in cheque processing by raising the automation threshold from the current ₹10,000 to ₹1 lakh. The move could bring more than 50–55% of the bank’s cheque volumes, by number, under automated processing, SBI Chief Information Officer Abhay Kishore Pandey said.
SBI currently uses image-based AI models to process cheques through its cheque truncation system. Cheques of up to ₹10,000 are already processed through an automated model, accounting for around 25% of the lender’s total cheque volumes.
The proposed increase in the threshold to ₹1 lakh is expected to substantially expand the share of transactions handled without conventional manual intervention. The bank expects the move to reduce processing costs while allowing employees involved in backend operations to focus more on customer-facing responsibilities.
AI to take cheque automation to the next level
SBI has already implemented straight-through processing for eligible cheques, with minimal human intervention in the transaction flow. The bank uses Vision Large Language Models to read cheque images, verify mandatory fields and check compliance requirements.
A control risk unit continues to review a sample of cheques processed through AI systems. The review helps identify errors or missed checks and allows the bank to determine whether its AI models require further training.
The next phase of automation is aimed at improving efficiency rather than simply replacing manual processes. SBI expects employees currently engaged in backend cheque processing to spend more time interacting with customers and supporting service delivery.
The bank is also making larger investments in enterprise-level AI infrastructure. Its annual technology expenditure is more than $2 billion, while the lender is increasing spending on AI platforms and related infrastructure with the expectation that the investment will generate lower operating costs over the longer term.
Same workforce, larger customer base
SBI’s broader AI strategy is focused on increasing the scale of services it can provide without a proportionate increase in its workforce. The bank expects automation to allow the same number of employees to serve a larger customer base while improving operational efficiency.
Another focus area is increasing the value generated from existing customers. SBI is exploring AI-driven customer engagement and analytics to improve customer lifetime value and encourage customers to maintain longer-term relationships with the bank.
The lender is also expanding the use of AI in lending. During FY26, SBI used AI-led underwriting to assess nearly ₹1 trillion in MSME loans of up to ₹5 crore each.
These underwriting models analyse GST information, credit bureau scores, account data and other structured and unstructured information to assess borrowers. The technology has also helped relationship managers by reducing the time required to collect information and conduct preliminary analysis.
AI enters credit decision-making
SBI has reported improvements in asset quality in portfolios assessed through its AI-enabled underwriting and business-rule systems. The bank has indicated that portfolios processed through these models recorded lower non-performing assets compared with relevant traditional processes.
AI has also become a regular subject of discussion at the board level. SBI’s IT Strategy Committee is examining investments and strategies for scaling AI across lending and other banking functions.
The growing number of customers seeking smaller loans is another reason the bank is accelerating AI-based credit decisioning. With loan demand increasingly coming from smaller-ticket borrowers rather than only large corporate customers, SBI believes faster and more efficient credit assessment will be essential.
The bank plans to combine Large Language Models and Small Language Models with internal data, external datasets and publicly available credit bureau information. The objective is to improve the speed and efficiency of lending decisions while maintaining the necessary risk and compliance checks.
SBI’s expanding AI strategy therefore extends beyond cheque processing. From automating routine banking operations to underwriting MSME loans and supporting faster credit decisions, the lender is positioning AI as a core part of its technology infrastructure, with cost optimisation, operational scalability and improved customer service among its key objectives.
