Vijay Shekhar Sharma Offers Personal Cheques to Founders Building AI Models

Paytm founder and CEO Vijay Shekhar Sharma is offering personal cheques to founders building artificial intelligence models, as he pushes for greater risk-taking in India’s emerging AI ecosystem. The move reflects Sharma’s increasingly direct involvement in AI beyond Paytm’s own technology strategy and comes as Indian founders and investors debate whether the country should build foundational models or concentrate on applications built on top of global AI systems.

Sharma has increasingly positioned AI as the next major technology opportunity for India. His latest approach is notable because it involves backing founders personally rather than relying only on Paytm’s corporate capital or its existing AI initiatives. The broader message is that India needs more entrepreneurs willing to take technically difficult bets in AI, even when the path to commercial returns is uncertain.

Key takeaways

• Vijay Shekhar Sharma is offering personal cheques to founders working on AI models.
• The move highlights his belief that India needs more risk capital for ambitious AI businesses.
• Sharma has previously argued that India should not surrender its AI opportunity to other countries.
• Paytm is separately building Paytm Intelligence, or Pi, as an AI-focused business.
• Paytm says AI could create new revenue lines and enterprise products beyond its traditional payments business.
• Sharma’s personal investing is distinct from Paytm’s corporate investment activities.
• The development comes amid increasing Indian investment in AI models, agents, infrastructure and specialised applications.

Why Sharma is backing AI founders personally

The most important part of Sharma’s latest move is not simply the size of any individual cheque. It is the willingness to put personal capital behind founders attempting to build AI technology.

Building AI models is significantly more capital-intensive and technically demanding than launching a conventional software startup. Founders may need access to expensive computing infrastructure, large datasets, specialised researchers and engineers, model-training expertise and substantial amounts of experimentation before a commercially useful product emerges.

That makes early-stage risk capital particularly important.

Sharma has argued that India has the talent and ecosystem required to build globally relevant AI businesses but needs entrepreneurs to take bigger bets. At the Global Fintech Fest, he said India should not allow its AI capabilities to become controlled by other countries and urged technology entrepreneurs to take the risk of building for the domestic market and beyond.

His decision to write personal cheques therefore fits into a broader position he has been expressing for several years: Indian entrepreneurs should not simply consume AI technology developed elsewhere.

Sharma’s AI thesis has evolved beyond Paytm

Sharma’s interest in AI is not new.

Paytm has been developing AI capabilities for years, initially focusing on areas such as fraud prevention, financial services and risk management. In 2023, Sharma said Paytm was building an India-scale AI system that could help financial institutions identify risks and fraud.

The strategy has since expanded.

At Paytm’s 26th annual general meeting in September 2026, Sharma described three categories of companies emerging in the AI era: businesses that use AI, companies that make AI a business, and companies that stay away from it. Paytm, he said, intends to both use AI internally and turn AI capabilities into a business.

That strategy has been brought together under Paytm Intelligence, or Pi.

The company says AI is already being used across engineering, merchant sales, loan collections, customer acquisition and retention. The next step is to commercialise some of those capabilities outside Paytm.

Paytm wants AI to become a revenue business

This distinction is important because Paytm’s AI strategy is no longer simply about cutting costs.

The company says it wants AI to generate entirely new revenue streams.

At its September AGM, Paytm said it was building systems that could help enterprises create an “agentic workforce” capable of handling workflows and business processes. Some of these services are already being piloted with enterprise customers.

Sharma expects AI revenue to begin appearing as identifiable business lines in roughly 500 days from the AGM, or about one and a half years. The company expects these revenues to emerge through its commerce and cloud businesses.

This creates an interesting connection between Sharma’s personal investments and Paytm’s corporate strategy.

His personal cheques can support the broader Indian AI ecosystem, while Paytm is attempting to build its own commercially viable AI products.

However, these should not be treated as the same pool of capital. A personal investment by Sharma does not automatically mean Paytm has invested in the startup or entered into a commercial relationship with it.

Why building AI models is different from building AI applications

India’s AI startup debate has increasingly divided into two broad approaches.

The first is building foundation models: large models capable of handling general-purpose tasks such as language, reasoning, coding, speech or multimodal processing.

The second is building applications and agents on top of existing models.

The first approach requires significantly greater infrastructure and capital.

A startup developing its own model may need to spend heavily before it has meaningful revenue. It also faces competition from companies with enormous computing resources and established research teams.

That makes personal backing from experienced technology entrepreneurs potentially valuable even when the cheque itself is relatively small.

Early capital can finance research teams, prototype models, data acquisition, compute usage and the first commercial experiments.

The bigger question is whether Indian startups can turn that initial technical advantage into defensible businesses.

Sharma has previously questioned India’s focus on foundation models

Interestingly, Sharma’s position has not always been that India must compete directly with the world’s largest AI laboratories.

In 2025, he argued that India should position itself as an “AI use case capital” rather than focusing exclusively on building large foundational models. He pointed to India’s scale and diverse economic environment as potential advantages for developing practical AI applications.

His more recent support for founders building models suggests a broader thesis may now be emerging.

India may need both layers: domestic model capabilities as well as businesses that turn AI into useful products.

This distinction matters because a country can become a major AI user without necessarily controlling the underlying technology. But building models without successful applications can also result in large amounts of capital being spent without sustainable commercial returns.

The strongest ecosystem would therefore combine research, models, infrastructure and applications.

Paytm is already positioning itself for that ecosystem

Paytm’s own AI ambitions provide another reason Sharma is closely watching the sector.

The company says its technology capabilities can be extended to small and medium-sized businesses across India. Rather than buying every AI capability externally, Paytm says it is developing native capabilities that can eventually become products.

That could eventually put Paytm into competition with enterprise AI companies, software providers and specialised AI-agent startups.

For small businesses, the opportunity is particularly significant.

A merchant could theoretically use an AI agent for customer service, sales, accounting, inventory, marketing or financial management without employing separate specialists for each function.

Sharma has previously described AI agents in similar terms, including the possibility of digital assistants performing functions comparable to a chief operating officer, chief financial officer or chief marketing officer for smaller businesses.

The AI investment environment in India is changing

The timing of Sharma’s personal investments is also significant.

Indian AI funding has moved well beyond experimentation. Startups are attracting capital across foundation models, enterprise AI, AI agents, developer tools, voice technology, cybersecurity and industry-specific applications.

Investors are increasingly looking for companies that can demonstrate either proprietary technology or a strong path to revenue.

Moneycontrol reported that AI startups had raised $1.56 billion across 206 deals through August 20, 2026, according to Venture Intelligence data. AI’s share of total deal value had risen to 23%, from 15% a year earlier.

That does not mean every AI startup will succeed.

The sector is still facing a fundamental capital-allocation question: how much money should be spent developing increasingly powerful models when global companies already provide sophisticated AI through APIs and open-weight releases?

Indian founders must therefore demonstrate why their models or infrastructure are differentiated.

Personal capital can send a different signal

A founder writing a personal cheque sends a different signal from a corporate investment.

Corporate venture capital generally involves investment committees, strategic objectives, financial return requirements and formal processes.

Personal investing can be more flexible.

An entrepreneur may decide to back a founder because of the team’s technical ability, ambition or insight before the startup has enough metrics to satisfy institutional investors.

That can be particularly useful in AI, where conventional startup metrics may not immediately capture the value of research progress.

At the same time, personal investment carries its own risks.

A founder-investor may have strong conviction about a technology that ultimately does not become commercially viable. AI model development can also require follow-on funding far beyond the initial cheque.

Therefore, Sharma’s backing should be viewed as a vote of confidence in the founders and the opportunity, rather than evidence that any particular AI model will succeed.

Sharma has been an active technology investor before

This is not Sharma’s first attempt to support technology startups with personal or investment capital.

In 2023, he launched VSS Investments, a SEBI-approved Category II alternative investment fund with an initial corpus of ₹20 crore and a ₹10 crore green-shoe option.

The fund was designed to invest in startups focused on areas including artificial intelligence and electric vehicles, particularly businesses serving Indian consumers and companies.

Sharma has also previously invested in a range of Indian technology and consumer startups.

His current interest in AI therefore represents an extension of an existing investing pattern rather than an entirely new activity.

What has changed is the strategic importance he assigns to AI.

Why India’s AI model race matters

The debate over domestic AI models is ultimately about more than startup valuations.

AI models are becoming a layer of digital infrastructure.

If Indian companies rely entirely on foreign models, they may have less control over pricing, availability, data policies and model behaviour. Domestic models could provide greater control for sensitive applications in areas such as finance, government, healthcare, defence and Indian-language services.

But developing those models domestically does not automatically create economic value.

Models need users.

They need developers building products on top of them. They need companies willing to pay for inference or software. And they need enough scale to justify the cost of computing infrastructure.

This is why the combination of model builders and application developers could be more important than either group alone.

The bigger picture

Sharma’s decision to offer personal cheques to AI-model founders is a small development in financial terms compared with the billions being spent globally on AI infrastructure. Its significance is more symbolic and strategic: an established Indian technology entrepreneur is willing to put his own money behind a new generation of AI builders.

For India’s startup ecosystem, that could help address one of the country’s recurring problems in deep technology: talented founders may have ideas and engineering capability but struggle to obtain sufficiently early risk capital.

Looking Ahead

The real test will be whether these early AI bets can progress from technical experimentation to durable businesses. India’s AI ecosystem is moving toward a more competitive phase in which founders will have to demonstrate not only model quality but also cost efficiency, proprietary data, distribution, enterprise adoption and a credible path to revenue.

For Sharma and Paytm, the AI opportunity is becoming broader than an internal technology upgrade. Paytm is building Pi as a potential new business while Sharma is personally encouraging founders to build AI capabilities in India. If both efforts succeed, they could give him exposure to the next generation of Indian AI businesses from both sides: as an operator and as an investor.