Daybreak - China's raising OpenClaw lobsters. India's testing the waters first

Episode Date: March 19, 2026

Every great Indian enterprise has a story worth telling in full.With The Intermission, The Ken's flagship long-form podcast, hosts Rohin Dharmakumar and Seetharaman G do exactly that—tracing the origins, turning points, and defining strategies of the companies that shaped modern India. Built on exhaustive research and proprietary interviews, each episode unfolds as a richly reported narrative. The first episode drops on March 23! Stay tuned!Last Friday, Razorpay CEO Harshil Mathur hosted 150 founders at Razorpay's Koramangala headquarters — not to talk payments but to let them showcase what they'd built with OpenClaw, the open-source AI agent taking the world by storm. The same week, thousands were queuing outside Baidu and Tencent offices in China just to get the software installed. The  open source agent AI platform is the same but the two approaches are quite different. China is deploying OpenClaw at a scale and speed no other country is matching. India, meanwhile, is moving carefully, deliberately, problem-first. So here's the question: is India behind China on OpenClaw? And is speed is the only thing that matters in the AI race?Read Inc42's report here: The New Garage: OpenClaw And India’s DIY AI Agent BoomDaybreak is produced from the newsroom of The Ken, India’s first subscriber-only business news platform. Subscribe for more exclusive, deeply-reported, and analytical business stories. 🚨The Ken's Zero Shot podcast is hosting a live event! This is a speculative yet realistic discussion built around one premise: what happens when AI agents take off in India? How will they rewire existing habits, business models and profit pools? Since nobody knows for sure, we won't pretend to have all the answers. Instead we are going to break the narrative. Click here for details. 

Hi, this is Rohan Dharma Kumar. If you've heard any of the Ken's podcast, you've probably heard me. My interruptions, my analogies and my contrarian takes on most topics. And you might rightly be wondering why am I interrupting this episode too. It's for a special announcement. For the last few months, I and Sita Raman Ganeshan, my colleague and the Ken's deputy editor, have been working on an ambitious new podcast. It's called Intermission.

We want to tell the secret sauce stories of India's greatest companies. Stories of how they were born, how they fought to survive, how they build their organizations and culture, how they manage to innovate and thrive over decades, and most importantly, how they're poised today. To do that, Sita and I have been reading books, poring over reports, going through financial statements, digging up archives, and talking to dozens of people. And if that wasn't enough, we also decided to throw in video into the mix. Yes, you heard that right. Intermission has also had to find its footing in the world of multi-camera shoots in professional studios, laborious editing, and extensive post-production.

Sita and I are still reeling from the intensity of our first studio recording. Intermission launches on March 23rd. To get alert, as soon as we release our first video. episode, please follow intermission on Spotify and Apple Podcast or subscribe to the Ken's YouTube channel. You can find all of the links at the ken.com slash I am. With that, back to your episode. Last Friday evening in Bangalore's Koremangla, about 150 founders showed up at Razorpe's headquarters. Razor Pay, as you know, is one of India's most consequential fintech companies. Over $9 billion in valuation.

and an IPO expected later this year. Its co-founder and CEO, Harshal Mathur, was the host of this event. But this gathering had nothing to do with payments. It was about OpenClaw, the open-source AI agent that has taken the global tech world by storm since January this year. Unlike a chatbot which waits for your question and then answers it, OpenClaught takes actions on your behalf, like managing emails, calendars, booking appointments, filling forms, all of it autonomously around the clock and running on your own machine. It was created by the Austrian developer Peter Steinberger and crossed 250,000 GitHub stars by March.

In fact, Steinberger has since been hired by OpenAI to build a next generation of personal agents. Now, here is what made that Friday evening in Coramangla matter. because the same week, something very different was happening in China. Thousands of people were queuing up outside Baidu's headquarters in Beijing and 10 cents in Shenzhen with laptops and mac minis, just to get someone to install the software. In fact, companies were forcing employees to prove their open-cler competency over the lunar new year under the threat of being fired.

Meanwhile, the Chinese government was issuing form. normal security warnings to state enterprises and banks against open claw. And this was happening as local governments were actively subsidizing startups to adopt the same tool. So here are two countries using the same open source agentic AI platform. And what's interesting is the way that each one has chosen to engage with it. It says something important about how they approach technological risk. And Harshal Mather's event goes on to. show exactly what this difference looks like.

When he announced the showcase on X, he was very clear about its purpose. He said this was a builder spotlight for founders who had already shipped something. And it would happen in front of a room that included OpenEI and Peak 15. His purpose was for India's builders to demonstrate what they had already made using the tool. So, is India behind China on OpenClaw? or is India being behind even the right way to look at this? Welcome to Daybreak, a business podcast from the Ken. I'm a host Nick Da Sharma and I don't chase the new cycle.

Instead, every day of the week, my colleague Rachel Vargheese and I will come to you with one business story that is worth understanding and worth your time. Today is Friday, the 20th of March. Let's start with China, because understanding the scale of what's happened there is essential to appreciate what makes India different. The Chinese have a phrase for it, Young Longshar, which means raising lobsters. It is after open-cloth's red crustacean logo,

and it's spread from developer forums to street slang to state media in the space of weeks, and the use cases span every level of society. At the top of the corporate ladder, Fuseng, who is the CEO of Chita Mobile, built a team of eight open-clothel. agents while recovering from a skiing injury. In 14 days, those agents were sending New Year greetings to over 600 contacts in four minutes,

publishing social media posts drawing millions of views and launching a website. All of this while Scheng was sleeping. And this story became the template, a one-person company running around the clock, compressing what used to take a team into a single autonomous system. The concept caught on every demographic. Wong Xiaoyan, a small business owner, told CNBC that she was using OpenClau specifically to build what the Chinese are now calling a one-person company or OPC. Human employees need rest, she said, but OpenClaw can run 24-7. Meanwhile, at a meetup in Beijing, a legal professional named Koki Shoe simply said, open claw has become really hot.

And at the corporate level, Bloomberg reported on a product manager at one of China's largest finance groups who was told that her team needed to stay back over the lunar new year to prove their open-claw competency with managers warning that anyone who did not demonstrate proficiency could be replaced. The contest was after all postponed after employee pushback, but the pressure did not ease. For everyday users, there were public install events. Baidu hosted hundreds in Beijing, Tencent organized a session outside its Chenjin headquarters that drew students, retirees, and office workers. One new user, Gongsheng, told CNBC, it seems everyone around me has it.

I don't want to be left behind. Apple's Mac minis sold out across the country because it had become the hardware of choice for running open-claw law. locally. The scale of this was so widespread that it triggered a government response. The central government of China issued formal warnings to state enterprises and banks. Employees at state-run financial institutions were told not to install open claw on office machines or personal phones connected to company networks. Those who already had it were asked to declare the installation for security review. China's cybersecurity agency published official guidelines,

and the People's Bank of China issued a separate advisory. But did this slow things down? The American cybersecurity firm's security scorecard says that China has already surpassed the United States in overall open-cloth adoption. And the install events are still running as of this week. Ryan Fedesuk at the American Enterprise Institute put it quite plainly. Banning Air Agents in 2026, he said, is like trying to ban spreadsheets in 1984. The productivity gains are way too visible and the cost of abstaining has become too obvious to ignore. But what the ban actually achieved was quite surgical in China.

It kept open claw away from sensitive state infrastructure while leaving the private sector and consumers free to continue. This tension, the central government restricting it on one hand, and local governments in Shenzhen and Wushi, offering subsidies of up to 2 million yuan to startups, building on it on the other is not confusion. As the diplomat very pertinently noted, it looks more like a deliberate calculation. Capture the economic upside of agentic AI while keeping it out of the party state's bloodstream. So what is India doing differently? And does that difference actually matter? Stay tuned to find out. So a lot of the details of the India's story and open clause use cases here come primarily from Inc.42's fantastic deep dive. It is the most detailed reporting that I have read

on this and I will link it in the show notes. Okay, so let me take you through three levels. The senior founder, the working professional and the small business owner because the contrast with China holds with each one of them. At the top, we have Harshal Mathur. But to understand why he was in that room on a Friday evening, you need to understand what agentic AI means to Razor Pay as a business. The same week as the OpenClawe showcase, Razor Pay launched Agent Studio at its FDX event in Delhi, a platform that lets businesses deploy AI agents directly on top of payment infrastructure,

automating dispute management, customer recovery, and payment reconciliation. His argument to business today was straightforward. He said, Razor Pay's ambition is to move from being a payment. provider to becoming a financial operating system for businesses. He also told A&I that every transaction on Razor Pay is now monitored by an AI agent scanning for fraud patterns in real time. So when Mathur says that he cares about open claw, the personal project and the professional agenda are the same thing.

The personal use that he shared with Inc 42 reflects exactly that sensibility. Using open claw with tightly restricted permissions, he built a Jarvis that reads his whooped sleep and recovery data, generates a morning briefing, orders food on Swiggy, auto sends gifts, and detects when he is in bed to turn the lights off. The most valuable insight actually came from correlating his Swiggy order history with his recovery metrics, what he ate and how it affected his sleep. His initial costs ran between $100 and $200 a day. So he implemented model routing,

reserving the expensive AI model for complex tasks only, and brought his daily spend down to between $30 and $50. Now, compare that to Fuseng's 600 New Year greetings in four minutes. Both of them are business leaders using the same tool. One is optimizing for output and the other for self-knowledge. And when Mathur hit the cost, wall, he studied it and fixed it himself. At the working professional level, a data scientist in Terra Data in Bangalore, reported by Inc.

42, built an open-claw agent for the union budget in February this year. A general news tracker would have been easy enough, but instead, he built one that pushed updates every 15 minutes, filtered specifically for Karnataka announcements and their impact on salaried professionals. When his API course nearly hit a thousand rupees in a single day, he did what Mathur did. He learned model routing and brought the cost down significantly. And at the small business level, and this is the most striking example in India, a software engineer named Raman Chaudhari left his Bangal job to help his dentist father.

The problem that he found was mundane, but genuine. Oral scanner companies like ITERO and 3-Shap each run separate portals. So a dental lab working with multiple dentists has to log into each one manually, download patient files, organize them and repeat it over and over again. Chaudhury used OpenClaw to automate that entire workflow and built a platform that he called Dent Node around it. There was no corporate mandate driving this, no government subsidy. and no competitive pressure.

This was a broken workflow in a dental lab in Jaipur, Rajasthan, and an engineer who understood the tool well enough fixed it. The pattern across all of these cases is consistent. A specific local problem, carefully scoped deployment, and when friction-like costs and the security risks appeared, it was studied and internalized rather than handed to someone else to clean up. Which brings us back to something that Harshal Mathur said at the India AI Impact Summit in February this year, that did not really get enough attention.

He drew a direct line between OpenClaw and UPI. He said, India took a leap with UPI and showed the world how real-time payments are supposed to be built. I think the same thing is going to happen with agent commerce. Now, this is a significant claim, but it has a foundation work. thinking about. UPI did not succeed because India deployed it faster than everybody else. It succeeded because the architecture was designed around a specific friction, and that was how to get hundreds of millions of people who have never had easy access to digital payments into the system. This took years of deliberate design before it was ready to scale, and when it did

scale, it reached a billion transactions a day. as we know is history. Now, China's pace of open claw adoption is extraordinary for sure, and so is the mobilization behind it. But there is a question worth thinking about. The Chinese users paying to install open claw and then paying again to remove it, learned something expensive. The engineer in Jaipur, who automated his father's dental lab,

learns something that will compound. These are two different types of progress. So if Mathur is right, if this really is India's UPI moment, then it will not be because India moved the fastest. It will be because when the technology was raw and the risk was real, the Indian builders chose to understand it rather than just deploy it. Daybreak is produced from the newsroom of the Ken, India's first subscriber-focused business news platform. What you're listening to is just a small sample of a subscriber-only offerings and a full subscription offers day. long-form feature stories, newsletters and a whole bunch of premium podcasts. To subscribe, head to the ken.com and click on the red subscribe button on the top of the website.

Today's episode was hosted and produced by my colleague Snitha Sharma and edited by Rajiv Sien.