The AI Shakeout, Delhi EV Policy 2.0 Draft, and Swiggy Cofounder Resigns

In the last three years, Indian AI looked unstoppable. Capital was flowing, founders were rebranding overnight, and every product with a chatbot suddenly felt investable. Then April 2026 arrived and exposed what a lot of people had been avoiding. India had raised roughly $10.5 billion into GenAI, but much of that money had gone into businesses that looked impressive in demos and weak in real life. Eighteen AI startups shut down, funding sentiment turned cold, and the market finally admitted something simple. Attention is easy to buy. Retention is not.

That is why this moment matters.

The real problem was not that India backed AI too early. It is that too much of the ecosystem built thin wrappers on top of Western models and mistook distribution for defensibility. A writing app, a legal drafting tool, a coding assistant, a summarizer. Many of them were effectively interfaces sitting on top of GPT or Claude. The moment those model providers shipped the same feature natively, the startup lost its reason to exist. What founders thought was a product often turned out to be a temporary layer in somebody else’s stack.

The user numbers made this worse. Indian GenAI apps could get downloads, curiosity, even buzz. But most of them could not hold users. Ninety percent of GenAI apps lost 60 percent of their monthly active users within 90 days. That is not a monetization problem. That is a product problem. People tried these tools, enjoyed the novelty, and then went back to a general assistant that could already do most of the job.

Enterprise AI was supposed to be safer. It wasn’t.

Indian enterprises were happy to run pilots, issue letters of intent, and talk about AI adoption in boardrooms. But when it came time to sign a large annual contract, hesitation kicked in. CIOs did not want to commit meaningful budgets to startups whose core intelligence still depended on US model providers. So startups hired teams, expanded burn, and waited for purchase orders that never came.

The companies that held up looked very different.

They were not building horizontal tools. They were solving narrow, high-friction problems. WorkOnGrid focused on utility systems like smart meters, billing, and loss detection. That is harder to sell, but harder to replace. Once embedded in workflows, AI stops being a feature and becomes infrastructure.

This is why the 2026 AI shakeout feels familiar. It resembles the SaaS correction of 2015, when generic CRM and ERP tools gave way to deeper platforms with real switching costs. The only difference is speed. That correction took years. This one took months.

There is also a talent constraint. Senior AI engineers now cost upwards of ₹70 lakh, making early hiring expensive. Junior roles are shrinking as AI automates entry-level work. At the same time, many legacy startups are rebranding as AI-first without real product change, and investors are starting to see through it.

So where does this leave Indian AI? Not dead. More serious.

The next phase is moving away from wrappers and toward infrastructure. Sovereign compute. Cooling systems. local data layers. vertical AI for sectors like utilities, lending, agriculture, and manufacturing. The winners are likely to be the companies building around Indian workflows, Indian regulation, and Indian data, not generic interfaces on top of foreign models.

Let’s go through what else is happening in Indian startup world - Grab your simmering cup of StartupChai.in and unwind with our hand-brewed memes.

“Do Sheeps Have Electric Dreams”: Delhi EV Policy 2.0 Draft

Delhi’s EV Policy 2.0 draft is going aggressive, offering zero road tax on electric cars priced up to ₹30 lakh while also proposing a ban on new non-electric two-wheeler registrations.

The move directly targets the city’s pollution crisis, where vehicles alone contribute nearly 23% of winter air pollution.

Read more here

“Man Changa Toh Kathauti Mein Ganga”: Karnataka Scraps Tax Breaks For EVs, Two-Wheelers Exempted

Karnataka has rolled back EV tax breaks for cars and other private vehicles, shifting to a price-based lifetime tax model, while keeping two-wheelers exempt.

The move comes even as EV adoption is picking up pace, partly driven by fuel uncertainties linked to the West Asian conflict.

Read more here

“Samsara Ka Niyam Hai Badlav”: Swiggy Cofounder Lakshmi Nandan Reddy Resigns, Three New Directors Appointed

Swiggy is seeing a boardroom reshuffle as cofounder Lakshmi Nandan Reddy Obul steps down to pursue other interests, alongside investor nominee Roger Clark Rabalais.

He’s been replaced by Renan De Castro Alves Pinto, while insiders tighten control with Phani Kishan Addepalli and Rahul Bothra joining as additional directors.

Read more here

“Suswagatam Suswagatam”: Vibhor Jain Appointed As MD & CEO Of ONDC

Vibhor Jain has officially been appointed MD and CEO of Open Network for Digital Commerce after serving in an acting role for the past year.

He steps into the role following the exit of founding CEO T Koshy, as the network looks to stabilise leadership.

Read more here

1. Swageazy has raised ₹5.4 Cr in a follow-on round led by Info Edge Ventures, with backing from founders of HR tech firms OnGrid and HROne. The round signals continued investor appetite for niche B2B gifting solutions.

   Read more here

2. BIDSO has raised ₹63 Cr in a Series A round led by Blume Ventures. The focus is clear, scaling design-led manufacturing for global consumer brands.

   Read more here