The numbers are hard to argue with; as of early 2026, more than 2.61 billion financial accounts are enabled on India’s Account Aggregator (AA) framework, with over 180 million accounts successfully linked. In the first half of FY26 alone, AAs facilitated an estimated Rs 1.47 lakh crore in loan disbursals across 15 million loans. Monthly volumes are now running at approximately Rs 24,000 crore, nearly double the previous half-year period. According to Sahamati’s H1 FY26 Impact Report, the AA framework now touches roughly 1 in 10 personal loans disbursed in India by value, accounting for over 10% of retail and MSME (micro, small and medium enterprise) lending by volume.
The rapid scaling of AAs reveals a complex structural challenge. For decades, Indian lending was stifled by information scarcity and a reliance on collateral. Today, the problem has inverted. Lenders are inundated with granular transaction data providing a real-time picture of a borrower’s financial life. Yet, more data does not inherently improve credit outcomes. Without sophisticated analytical layers, an abundance of data leads to cognitive overload or algorithmic noise that masks underlying risks.
The mirage of data volume
For most of India’s lending history, the primary constraint was visibility. Large sections of the borrowing population, including informal workers and first-time borrowers, remained invisible to traditional underwriting. The credit gap for MSMEs, according to 2025 SIDBI research, is currently estimated at close to Rs 30 lakh crore, with the deficit widest among micro-enterprises and the rural self-employed. The AA framework addresses this by bringing verified, structured financial data into lender workflows through time-bound and purpose-bound consent.
The challenge is that data access and analytical capability do not scale at the same rate. A lender receiving six months of bank statement data through an AA journey has far greater visibility than was previously possible, but raw transaction data does not interpret itself. A large credit inflow may reflect a business receipt, a family transfer, or a seasonal fluctuation. High account balances on a particular date might conceal weeks of liquidity stress. The meaning of any individual signal depends entirely on context, requiring analytical frameworks that many lenders have yet to fully mature.
This is not a theoretical concern. Over 68% of MSMEs seeking credit in the past year were new-to-credit. For these borrowers, transactional data is often the only underwriting input available. If misread, the consequences are real: eligible borrowers are denied credit, while ineligible ones are approved at mispriced rates, leading to portfolio stress that only surfaces quarters later.
Building the intelligence layer
The dominant narrative in open finance often conflates data availability with utility. In practice, data abundance creates its own pathologies. Models trained on noisy signals develop false confidence, and approval decisions that appear data-driven may actually replicate structural biases. Risk is frequently mispriced not because lenders lack information, but because they lack systems to distinguish meaningful patterns from incidental ones.
Building a robust intelligence layer requires disciplined transaction categorisation to separate business income from personal inflows. It demands cash flow modelling that accounts for income seasonality and sector-specific rhythms. It also requires anomaly detection that identifies data manipulation without penalising legitimate financial irregularity. Additionally, scoring frameworks must be explainable to both regulators and borrowers to remain compliant with the evolving Digital Personal Data Protection (DPDP) landscape.
The AA model is built around granular permissions. This is a design philosophy reflecting the view that data-sharing frameworks should expand borrower agency, not simply reduce friction for institutional data collection. This philosophy must carry forward into how lenders build their analytical frameworks.
Where competitive advantage lies
The next phase of India’s open finance journey will not be won on data access. What will differentiate institutions is the quality of interpretation: the ability to extract genuine signals from financial data and understand a borrower’s actual cash flow position rather than a point-in-time balance. Lenders who treat AA data as merely a faster version of a traditional document will not realise the potential of the framework. Those who invest in the analytical infrastructure to contextualise this data will lead the market.
The AA framework has effectively addressed the structural challenge of information scarcity by establishing interoperable, consent-based infrastructure at a national scale. However, the objective of these rails was never merely to maximise the volume of data in motion. Instead, the goal is to create conditions where superior information leads to superior outcomes: faster decisions, fairer pricing, and expanded financial access. In this new era, competitive advantage will belong to those with the most disciplined analytical frameworks. Data portability is a means to an end, not the end itself. The success of open finance will ultimately be measured not by the complexity of the rails, but by the quality of the judgment used to navigate them.
Author is Co-Founder and CEO, Finarkein. Views are personal.
