As AI moves faster than ever before

AI Summit - India – 16th to 21st Feb 2026 at Bharat Mandapam- New Delhi - AI and Beyond

In February 2026, Bharat Mandapam, India’s largest exhibition and convention ground in New Delhi, which was inaugurated in 2023 and previously hosted the G20 summit, echoed India’s transition to a Sovereign Digital India. Since the emergence of OpenAI in late 2022, it has expanded across domains, from initial handling of coding and chat-related queries to processing multi-step workflow tasks with agentic AI, and now to reasoning tasks as well. The entire AI ecosystem has evolved tremendously.

The growth of AI introduced safety risks, job losses as workflows became autonomous, and, finally, the question of the AI bubble was doing the rounds, with 95% of AI systems not achieving the desired result and 55% of the business community stating that they regret the decision to replace humans with AI and are ready to look back on their decision. Despite all the factors preceding the event and creating headwinds, the India AI summit highlighted the push to move beyond AI risks and make it part of a democratised society, where countries now start accepting AI as the way forward, with the future of AI shifting directly to laptops, smartphones and eye-gears, placing a stress on data autonomy through localised models, improving the contribution to other sectors, and ensuring safety and dependence.

The MANAV framework introduced by the Prime Minister of India was a step towards a more humanised process for ‘M- Moral systems’, ‘A-Accountable Governance’, ‘N- National Sovereignty’, ‘A- Assessable AI’ and ‘V-Valid Operating System’ in the view of developing a Sovereign Digital India landscape. 

‘Sarvajana Hitaya, Sarvan Sukhaya !’, as it was themed, was a digital push by the Indian Prime Minister, Mr Narender Modi, towards making India future-ready. How it shapes up remains to be seen, but the episode marked India's presence on the AI world's diaspora map.

The Indian AI Summit 2026, which concluded on 21st Feb 2026 in New Delhi, was a first-of-its-kind event held in the Global South. The message is clear. India can no longer be ignored in the new world equation and is positioning itself as a strong geopolitical powerhouse in the AI narrative. India, already ranked 3rd in the 2025 global AI vibrancy tool report by Sanford University with a 21.59 score, just after the USA (78.6) and China (36.95), is trying to close the AI growth gap, which is highly concentrated in developed nations, with major world corporations focusing on future AI headquartered in the USA. Before the Summit, India had only a handful of business houses featured in the AI ecosystem. The platform successfully showcased India's readiness, with its 600+ AI startups and technology penetration already making headway worldwide through innovations like ‘Chewie’, a wet kitchen waste dispenser priced at 37900/- by a Bangalore-based start-up, already pre-booked for production till 2027.

Broad Takeaways From The Summit: A Big Snarl That The Future Will Remember

1. The summit featured technology hubs focused on AI safety rather than risk alone, with many podiums dedicated to this work and to innovation-linked startups. The need to shift the focus from theoretical risks and limitations to real-world AI applications for their advantage was clear.
2. The India AI summit also delivered a strong message that the focus should be on both creating new AI models and Applied AI - using the technology in developing sectors such as health, agriculture and education. The focus now shifts from just AI risk to how AI can enhance the day-to-day work process.
3. India is playing a key role in bridging AI know-how between the superpowers and developing nations in South Asia.
4. India has given the go-ahead to the semiconductor supply chain coalition led by the USA (Pax Silica).
5. India is registering its stance that it wants to be taken seriously by foreign tech giants, not merely used as a data centre, through initiatives like Bharat Gen.
6. The summit also saw pledges of more than $250 billion in investment in AI infrastructure, led by Reliance Jio ($110B), Adani Group ($100B), Microsoft ($50B), and Google ($15B).
7. India’s indigenously built AI Sarvam, asserting India’s way forward in building ‘Make in India AI models’
8. India’s preparedness to host such a mega event at the global level was successful, but it was also significantly challenged by the need for several improvements.
9. The strong global participation, with major industry giants in attendance, shows the staggering impact India has on the global stage.
10. The government’s addition of 20000 GPUs to the existing limit, stamping the central government's push for an AI future in India and participation.

Summarising some of the key Indian contributions witnessed firsthand by the team of Academic Mantra services at the AI summit, which were buried amid the galore of controversy and the kiosks and podiums of global giants.

1. Chewie: The first on the list is the Kitchen waste dispenser by Mankomb Technologies in Bangalore, founded in Jul 2024 by Mrudul, Dinesh and Jawahar. The company already has its order book filled for 2027, based on its current capacity, which was a key highlight of the event. The product can process 120kg of waste per month, uses AI to identify the composition of the waste it receives, and produces composite waste within 48 hrs. From meat bones to peels and cooked kitchen scraps, it can process all of them into nutrient-rich soil. It's compact and creates odour-free, pest-free composites, eliminating the entire hassle of the indoor process of composite creation. 

2. Tyrone: Netweb and Tyrone systems, which have built a series of supercomputing GPU systems powering Indian data centres, are planning to scale up a new high-end portfolio with the NVIDIA GB10 chip. They have developed liquid-cooling devices that eliminate the need for fan-based cooling, optimising space and allowing them to fit more than 10 GPUs in a single system, making it one of the most powerful systems in the world.

3. Intellect Purple Fabric: The product Enterprise AI on tap is developed by an Indian company, Intellect Design Arena, that was in the information technology business and, with the advent of AI, has now started developing KRC products: knowledge, reasoning, context-based. The product offered 150 domain-trained AI agents to help generate faster, reliable results. They guarantee 95% output accuracy for their product, thereby exceeding the average accuracy of AI models. They have designed the product for SMEs like education, finance, etc. Their pricing for the base model starts at 99500 for an average employee size of up to 50. The product use case scenarios are for businesses that want help with production planning, market lead automation, or corporate pitching.

4. Vecros: Another Bangalore company was showcasing a drone docking station it had developed. The dock supports drones that can fly even without GPS. The company was promoting India’s first fully built AI-driven drone carrier, offering self-driving drones without a pilot, powered by a dock-based ecosystem that allows the drone to self-change or recharge its battery without human intervention. The drone can fly at night and has JetCore chips installed. They also allowed the drones to self-station themselves at the dock station in case of heavy rain, operate indigenously to perform tasks, and are waterproof. 

5. Lex Legis: A company focused on driving transformation in legal platforms through AI. The company has planned an intervention with 4 impact solutions to ensure maximum security and legal compliance. The company plans to offer case research for lawyers and government legal departments where files are stuck due to pendency and a lack of reference case studies. They aim to mitigate the risk that an AI search now generates a case study based on a user's search and request, even if it’s not actual. They are powered by NVIDIA 3 Nano for accuracy. Thus, offering legitimate case studies at users' fingertips is a vital solution they are banking on. The 3rd major ease of change they are offering is the drafting of legal documents, such as bail applications and other legal documents that require you to attend a legal advisory. The fourth service they are engaged in is sorting pending data for government departments by priority for tracking and closures. As of date, there are 4 lac cases, as they say, which will make it easier for the government to discharge them. The service is not available to retail users; they plan to introduce it at a later stage. They have now partnered with L&T Vyoma to ensure data privacy and also store data locally in India.

6. Rudra: The next on the list that caught our eye was the Rudra 3 Server – the first Indian-designed server, with most of the construction done in Kochi. They currently use an NVIDIA GPU for heavy workloads and LLM models. They have developed their own chassis and offer research services to PhD students who require high-quality research. They use Rawat's cloud services. They also offer server services to government departments. They fall under the Ministry of Intelligence, where access to research documents is key. They are fast and provide reliable services using a 20-bit processor unit, supporting the healthcare, manufacturing, and weather industries. Top institutions have already installed the devices for AI research.

7. The seventh on the list is Sarvam:The Bangalore-based company launched two models, Sarvam ‘30 B’ and ‘105 B’, focused on 22 languages, and that work even on basic feature mobiles. CEO Vivek Raghavan highlighted that they have hit the right chord with the Indian community with this approach, as they have identified a critical gap that other AI products from the West have failed to address. For Indians, service in the local language is key to broadening your reach, as dialects change on average every 15-100 kilometres you travel

   The company is launching smart glasses called ‘Kaze’ by May 26. It’s another indigenous product, built for mass use at an affordable price and focused on localised use with real-time voice assistance. For example, farmers can point to their crops and get AI assistance on their crop health. A shopkeeper can also now get real-time updates on their inventory. Compared with foreign products, which have already seen billions of dollars invested by Meta, partnered with Ray-Ban, and cost more than Rs 30000, the Sarvam smart glasses work in 10 languages, including English.

Amidst the hustle and bustle of some fake narratives, people and companies playing for quick limelight and gain, there was one Robo dog – ‘Svan M2’ – which was actually indigenous but buried under the continuous stream of negative news. An industrial quadruped robot built by IIT Kanpur and Xtera Robotics was showcased by L&T as an industrial robot designed for construction site monitoring. The robot could carry a payload of 5 kg and was equipped with LiDAR and cameras for 3D mapping, thermal imaging, and leak detection. The robot could be seen climbing stairs and navigating rough terrain, typical of construction sites.

Although the summit was a grand premiere for India’s AI future, a few concerns remain to be addressed:

1. Logistics: Hosting a global event of this scale, with the world’s largest population, is not going to be easy for any country. The AI summit app and entry, facilitated by a ‘Digi card’ that provided easy access to the venue, were good initiatives. However, hosting the event in the capital, which faces transportation blockades, heavy traffic, congestion, and security issues, remained a challenge due to crowd mismanagement, even at the speaker podium where the keynotes were delivered. The 1st day saw major crowd management hiccups, and the throng of VIP delegates descending on the platform caused significant discomfort for attendees and even exhibitors.

2. The robodog Orion, copied by the Golgotia University, claimed that a Chinese robot, which was actually bought by them, was made by them in their lab, causing serious damage to a country that was tabling itself as a strong AI player. 

Although online viewership figures are not flagged, a widely publicised event that had the world’s media covering the content live for six days received a mixed response from the community at large. Barring a few disappointments, the AI event was largely a success, with 5 lakh-plus attendees, 118+ country representatives, 20 state heads, 3000 speakers, business leaders, and stalwarts such as Sam Altman, Sundar Pichai, Yan Le Chun, and Yoshua Bengio, who graced the occasion, making it a highly talked-about event in recent days. The AI summit also saw more than 2 lakh young people attend on the first day, accelerating AI adoption among youth. The shift from AI risk to pivoting towards AI adoption was also a key factor for the summit in the AI future roadmap.