Many artificial intelligence startup ideas remain superficial “wrappers” built on top of existing models. But as AI model makers add more capabilities, investors are wary of startups that could easily become redundant.
Case in point: After reviewing more than 4,000 applications for India's joint AI accelerator for startups run by Google and venture firm Accel, “wrapper” ideas dominated. However, Accel partner Prayank Swaroop told TechCrunch that none of them were among the five startups in the latest cohort (pictured above).
Announced in November, the AI-focused Atoms program by Google and Accel aims to support early-stage startups building AI products relevant to India. Startups selected for the latest cohort will receive up to $2 million in funding from Accel and Google's AI Futures Fund, as well as up to $350,000 in cloud and AI computing credits from Google, the companies said.
About 70% of the rejected applications were “wrappers,” or startups that layered AI capabilities such as chatbots on top of existing software, but “didn't use AI to reimagine new workflows,” Swaroop said.
Swaroop said many of the remaining rejected applications fell into crowded categories, such as marketing automation and AI recruiting tools, in areas where investors see little novelty. Startups in these areas often struggle to differentiate, he said.
Perhaps this is not surprising. This year's program received nearly four times as many applications as previous Accel's Atoms cohorts and included many first-time founders.
India's growing AI ecosystem remains primarily focused on enterprise applications, and Swaroop said the applications reflect that. Approximately 62% of the entries focused on productivity tools, and an additional 13% focused on software development and coding. This means that about three-quarters of the applications were enterprise software ideas rather than consumer products. (Swaroop was looking forward to more ideas on health care and education.)
tech crunch event
San Francisco, CA | October 13-15, 2026
Jonathan Silver, co-founder and director of the Google AI Futures Fund, said the five startups selected closely align with areas where Google expects AI to be more deeply deployed in the real world.
Silver said the program does not require startups to use Google's model exclusively, noting that many companies combine multiple models depending on their workflow. The goal, he said, is to gather feedback from startups about how Google's model works in real-world applications.
Insights from these startups are fed back to the Google DeepMind team to help improve future models, creating what Silver described as a “flywheel” between startup experimentation and AI development. “The fact that companies are using alternative models means that Google has work to do to build the best model on the market,” he told TechCrunch.
This year's selected startups are:
K-Dense is building AI “co-scientists” to accelerate research in areas such as life sciences and chemistry. Dodge.ai develops autonomous agents for enterprise ERP systems. Persistence Labs focuses on voice AI for call center operations. Zingroll is building a platform for AI-generated movies and shows. Level Plane applies AI to industrial automation in automotive and aerospace manufacturing.
Source link