Why Accel and Google are chasing India’s newest AI founders

The target startups must be a part of Accel's Atoms accelerator programme - an early-stage support initiative for Indian and Indian-origin founders. Accel started the Atoms programme back in 2021 and has so far invested in nine AI companies through the accelerator. These are: Anko, BPR Hub, dubdub.ai, FinSynth, HighRev.ai, Ripik.ai, Rocket (formerly known as DhiWise), Skoob, and Zingle.

Since the launch of Atoms, Accel has seen a shift in the kind of AI companies it backs. “India has definitely been a hotbed for both consumer and enterprise companies. The challenge on the consumer side is that Indians are very cost-conscious," said Prayank Swaroop, partner at Accel, at a virtual media roundtable.

“But we are beginning to see a lot of India-specific innovation. Now that model prices are coming down, we want founders to build more consumer solutions for the same Indian consumers."

Accel is now looking to back companies across four broad sectors as part of its 2026 AI cohort, which begins in February. These are coding (Future of Coding), productivity (Future of Productivity), creativity (Future of Creativity), and entertainment (Future of Entertainment).

Earlier this year, Accel invested in two consumer AI startups. Along with Prosus, it backed Arivihan, a local language online learning platform, which raised $4.17 million in a seed round, and healthcare AI assistant August AI, which raised $3 million.

For AIFF, the choice to back Indian AI startups was deliberate. “We directly want to back the country's most promising AI talent and really fuel that next wave of AI innovation that's going to come from India for the rest of the world," said Jonathan Silber, co-founder and director of AIFF.

Apart from the equity investments that the two firms are making, startups that are part of the new cohort will also get Google Cloud credits worth $350,000. Startups can use these credits for Google's other offerings, too, such as Gemini AI models, and AI research laboratory DeepMind. “When we find interesting companies that are a good fit, we pull in our own folks. Our teams can receive the model feedback from our portfolio companies so that they can make the business better," said Silber. “They can improve the product. They can figure out what features to incorporated into the road map. So, it's kind of a give-and-take on both sides."

Large AI companies from the US have increasingly been setting up operations in India, primarily led by the size of the country's population. "The volume of people that India has, you won't get access to anywhere else in the world, bar China. Secondly, the amount of languages that India has. But overall, the quantum of data is so high that models become more intelligent with more access to data and can then give better answers," said Raja Lahiri, partner and technology industry leader at Grant Thornton Bharat LLP.

Such co-investments also mean that more AI-first startups enter the ecosystem, which means that more innovative solutions can be developed. "Companies can't automate everything in one day, but they can pick different silos, like HR, recruitment finance, things like that to build and automate. Startups that are forming to solve these problems are what are driving venture capital investments," said Lahiri.

Notably, Accel's partnership with AIFF is the second of its kind for Atoms. In October, the firm partnered with tech investor Prosus to co-invest $1 million each in what it calls “leaptech" companies, as part of the Atoms X track of its accelerator programme.

"Partnerships between venture funds and global deep-tech companies could play an increasingly important role in strengthening the early-stage ecosystem. By bringing together capital and potentially cutting-edge technology, they give young startups a stronger foundation to innovate and scale sustainably," said Amit Nawka, partner, deals, at PwC India.

Both Accel and AIFF declined to state how much of equity they'd be picking up in the companies that were going to be part of their accelerator programmes. “Globally, we've done investments where we have single-digit stakes because we just believe the company will do phenomenally well," said Swaroop. “Traditionally, I think whatever the India market is pricing the company at or if they are attracting international VCs, then we will have to compete against those terms."

Acquisition or long-term bets?

Earlier this year, Google announced one of its largest acquisitions to date, a $32-billion buyout of Israeli cloud security company Wiz. The deal is pending regulatory approval.

Similarly, it was on track to acquire AI coding assistant Windsurf, but ended up in a $2.4 billion deal to hire key executives as well as access to some of the company's technology on non-exclusive terms.

While picking up early, big equity stakes may look like a way to ease future AI acquisitions, Silber said that isn’t the intent—though he didn’t rule out the possibility.

“If at some point down the line, one of these companies makes sense for Google to acquire, it could be looked at, but that's really not at all the intention of how we're building this investment entity."

According to a report from M&A platform Dealroom, AI venture capital investments hit an all-time high in 2024, reaching $110 billion, up 62% year-on-year. Early-stage investor Antler had the most deals at 74, followed by Andreessen Horowitz at 66 and General Catalyst at 55. At 34 deals, Accel was ranked eighth.