A string of startup shutdowns and strategic pivots in the artificial intelligence space is prompting venture capital (VC) firms to become more selective, with investors increasingly moving away from generic AI offerings and focusing instead on businesses with stronger differentiation and clearer commercial outcomes.
Investors say the shift reflects a broader reassessment of the AI startup landscape after a period of aggressive funding during 2023 and 2024, when many companies raised capital on expectations of rapid growth and durable advantages that have proved difficult to sustain.
According to Tracxn data, 18 AI startups in India have shut down in the last 16 months. Some of the companies that have ceased operations include Builder AI, Neuropixel, Dotagent, Zeda io, LEGOAI and MoneyyAI. Even Ola’s AI venture Krutrim has reportedly pulled back its consumer assistant Kruti from app stores and the web as it pivots towards AI cloud infrastructure.
Wrapper Collapse
For investors, the shutdowns have reinforced concerns around business durability in a market where capabilities of large technology firms and foundation model providers continue to expand rapidly.
“A big part of our caution comes from the fact that large global technology and LLM players have a very high interest in building horizontal AI capabilities into their own stack,” Satyakam Mohanty, co-founder and managing partner at Wyser Capital, said. “So the moat you think you have can shrink quickly when one of them decides to move into your space.”
The concern is particularly acute for startups that build products as a thin layer on top of existing foundation models without underlying intellectual property or deeper customer integration. Investors say advances in the underlying technology stack can quickly reduce the relevance of such offerings.
VC firms are therefore showing greater preference for applied AI models where technology solves specific business problems and becomes embedded into customer workflows.
“Our preference is for applied AI, where technology directly improves productivity, reduces costs or creates measurable efficiency gains across manufacturing, healthcare, fintech infrastructure and enterprise workflows,” said Apoorva Ranjan Sharma, co-founder of Venture Catalysts.
Retooling the Pitch
Industry executives say another challenge has been the gap between experimentation and actual business adoption. Many AI startups have secured pilot projects but have struggled to convert them into recurring commercial contracts.
“Very quickly, AI startups are benchmarked against the best global companies for speed of iteration and scaling, which is a tall order. Also, lots of pilots are not converting into real commercial traction,” Deepak Gupta, general partner at WEH Ventures, said.
VC firms say founder behaviour is also changing. Instead of building around models first, entrepreneurs are spending more time validating use cases, distribution and operational integration before taking products to market.
“The founders we are most drawn to right now are the ones who have spent real time on the operational or distribution layer before spending time on the model itself,” said Maanav Sagar, managing partner at Good Capital.
The shift suggests that after an initial AI funding rush, investors are increasingly backing businesses where the advantage lies beyond the model itself.