Banks are moving beyond buying AI tools, becoming investors, strategic partners and co-builders with AI startups, particularly with an eye on creating AI-powered experiences for their customers
As AI moves deeper into critical banking workflows, institutions want greater control, customisation and influence over the technology they depend on
After a wave of global banks backing AI labs in the West, India is also beginning to see this shift with HDFC Bank and Canara Bank’s CoRover bet and IDFC FIRST Bank’s collaboration with Sarvam AI
Rogo, a US-based financial-services AI startup, raised $30 Mn this month. What made this unique was that some of the world’s largest banks, including Barclays, BNP Paribas, Citi, MUFG and Société Générale, invested in this company, even though it has VC backing too.
Nine global banks are now on Rogo’s cap table, representing nearly $20 Tn in combined assets.
Rogo isn’t an isolated case. Large banks and their venture investment arms are increasingly taking big bets on AI companies.
Such investments have grown at a compounded annual rate of 21% since 2023, per an Evident AI analysis of the 50 biggest global banks, with Wells Fargo, Citi and Goldman Sachs the most active US bank investors.
Recent deals around the world and in India back this trend. Citigroup and Spain’s Santander have both taken stakes in Tokyo-based Sakana AI, which builds foundational models and custom AI for financial services.
In India, examples include HDFC Bank and Canara Bank’s investments in CoRover as well as IDFC Bank’s partnership with Sarvam to build an R&D lab with the aim of developing the world’s first self-improving bank.
Three forces seem to be pushing banks past the old vendor relationship. Deep integration into regulated, high-stakes workflows requires more trust and customisation than an arm’s-length license can provide.
Owning a stake or co-building the R&D gives banks influence over a roadmap they’re increasingly dependent on. And for BFSI institutions sitting on enormous proprietary data and distribution, contributing that data to jointly built AI systems can be more valuable than a services contract.
“License provides a means to access technology, while investment or co-development may result in a more profound collaboration between a bank and a startup. Cocreation gives the bank an ability to affect product architecture and work flows instead of simply fitting the bank’s processes into a ready-made solution,” said Mukesh Pandey, founder of consultancy firm Rupyaa Paisa.
Whether this becomes the default model for bank-AI relationships, or stays concentrated among the largest, most AI-forward institutions, is still an open question.
So is what this means for smaller AI startups negotiating with a bank that is simultaneously their client, their investor and increasingly their co-developer. The ones that end up without any bank partnerships or strategic investments may struggle to get the same access to data, distribution and trust.
HDFC Bank’s move on CoRover, the Bengaluru-based conversational AI firm behind BharatGPT, was described as the first investment by a leading private bank in a model maker.
That isn’t CoRover’s only investor with a bank. Canbank Venture Capital Fund, the wholly owned VC arm of Canara Bank, had already invested in CoRover.
CoRover’s founder and CEO Ankush Sabhrawal said that the association between the company and the banking ecosystem has moved from being a typical startup-investor partnership to a strong technological and domain collaboration.
Further, Sarvam’s partnership with IDFC FIRST Bank is a particularly interesting example. The companies said they would work towards building what they describe as a “self-improving bank.”
Their collaborative R&D lab will focus on frontier AI research, post-training and AI safety, with systems designed to learn from customer interactions, human corrections, exceptions and eventual outcomes. The partnership is intended to bring AI research closer to the bank’s own operating environment, allowing models and systems to be tested against the institution’s specific problems and improved using production outcomes.
Sarvam has also built a growing footprint across financial services through more deployment-oriented partnerships. With Mahindra Finance, the startup’s voice AI has been used across customer-facing workflows, and Sarvam’s current enterprise portfolio lists Mahindra Finance alongside IDFC FIRST Bank, Tata Capital and SBI Life among its financial-services customers.
While not a bank, Bajaj Finance recently picked up a 5% stake in Delhi NCR-based TrueFan AI, a generative-video startup whose avatars Bajaj was already using at scale for customer engagement and dealer communication. The deal sits under the NBFC’s new Finserv Intelligence programme, which plans to invest ₹1,500–2,000 Cr into early stage AI, cybersecurity and fintech startups.
For banks and indeed NBFCs, AI is increasingly becoming more than another technology procurement category. Besides just being mere customers, banks and NBFCs are becoming financial stakeholders in the company building the technology.
India’s banking and financial-services sector is beginning to show early signs of the same shift.
Beyond financial resources, bank-led investments bring with them insights into the operational nature of financial services, including areas where AI technologies could be applied effectively in the area of customer support, employee assistance, knowledge management, multilingual engagement, etc.
“The feedback provided by banks becomes useful for determining specific areas where AI needs more accuracy, context, security, and compliance. It is crucial for BharatGPT, as the majority of banking transactions involve multilingual clients and difficult financial terms. The company views the controlled customization as a use of the client’s approved knowledge reference points, algorithms, and mechanisms instead of utilising proprietary information in a standard solution,” CoRover’s Sabharwal added.
That kind of close, iterative feedback loop is easier to justify when the sector on the other end of it is already AI’s biggest spender.
BFSI holds the largest share of India’s AI market, largely on the back of early adoption for fraud detection, risk management, customer service, and personalised financial solutions. The sector’s focus on digital transformation and data-driven decision-making has made it a key driver of AI demand — reaching a market size of over $1 Bn in 2025, and projected by research firm IMARC Group to grow to $9.4 Bn by 2034 at a CAGR of 26.5%.
A more significant change may actually be happening outside the cap table. Indian banks and financial institutions are increasingly working directly with AI startups to adapt technology to specific banking workflows, rather than simply purchasing a general-purpose AI product.
The logic behind these tie-ups was laid out plainly by Axis Bank’s Prasad Lad, the head of the bank’s business intelligence unit.
He said Axis is working with startups across voice AI, document reading and rapid UI/UX prototyping, adding that this isn’t an area where the bank can build everything in-house. “Startups such as Sarvam bring capabilities, particularly across Indian languages, that would be difficult for a bank to develop internally,” he added.
Kotak Mahindra Bank, on the other hand, has started leveraging AI across software development, employee productivity, customer service and experience, credit decisioning, and marketing. While most of the AI work is done by in-house team, the bank leans on startup partners for niche requirements, said Vijay Narayanan, Head of Innovation and AI, without specifying the startups Kotak Mahindra Bank is currently working with.
That same efficiency the AI partnerships are built to deliver is already showing up in headcount. Axis Bank reported a decline of over 3,100 employees in FY26, which it attributed to sustained investments in technology and productivity enhancements. HDFC Bank, Kotak Mahindra Bank and ICICI Bank have seen similar headcount reductions, directly or indirectly tied to AI-led automation.
In some cases, roles are being revised. “We expect 50-60% of our employees to be repurposed into sales and relationship roles over the next two years, with no layoffs as part of this transition but we continue to hire, especially for specialised technology talent,” Salee S Nair, MD and CEO, Tamil Nadu Small Mercantile Bank.
As with other layoffs in recent times, many have wondered whether these strategic investments are resulting in less humans in the banking loop.
Axis Bank CEO Amitabh Chaudhry told Bloomberg that investments in AI and technology would let the bank expand its business without a proportional rise in staff, signalling slower hiring even as operations scale up.
This is not a new experience for Indian banks by any means.
Every step taken by Indian banks in the past two decades has come with such cutbanks — whether it is internet banking, mobile banking and digital payments or indeed the AI banking future that banks are rushing to embrace.
But with every new technology jump, there have been teething pains, bugs and major tech gaps, which is why the buck always stops at the humans leading the banks. With AI, this human responsibility will grow by magnitudes because fewer humans will be running the banks.
So even if banks are ready for an AI future as they are demonstrating today with their investments, the next major discussion has to cover what the role of humans will be in this future.
Source: Inc42
