{"id":89846,"date":"2026-02-15T11:14:09","date_gmt":"2026-02-15T14:14:09","guid":{"rendered":"https:\/\/tech.einnews.com\/article\/892561659"},"modified":"2026-02-15T11:14:09","modified_gmt":"2026-02-15T14:14:09","slug":"tech-mahindra-bets-on-world-models-and-rd-for-ai-race","status":"publish","type":"post","link":"https:\/\/new7.shop\/zerocostfreehost\/index.php\/2026\/02\/15\/tech-mahindra-bets-on-world-models-and-rd-for-ai-race\/","title":{"rendered":"Tech Mahindra bets on world models and R&D for AI race"},"content":{"rendered":"
<\/div>\n

India\u2019s IT company Tech Mahindra has begun building its own AI world models, capable of simulating real world scenarios for sectors like telecom and healthcare, in its bid to stay ahead of the AI curve in India.<\/p>\n

\u201cAt this point in time customers are not asking for [such models], but in six months to one year they are going to ask that the AI be contextually relevant to the change that appears in a process. Attention to such levers allows Tech Mahindra to plan future actions,\u201d said Nikhil Malhotra, Chief Innovation Officer and Global Head of AI and Emerging Tech, Tech Mahindra.<\/p>\n

Currently, 70 per cent of Tech Mahindra\u2019s new, large deals (TCV of $1 billion as per Q3 earnings) have AI embedded into them. It is also delving into the problems of security with the advent of Anthropic\u2019s MCP protocol, and test how MCP security needs to be done with the agentic enterprise.<\/p>\n

Pitfalls of Gen AI<\/h2>\n

\u201cWhen the world was talking about Gen AI, we were already talking about agents and the pitfalls of Gen AI, which was responsibility. Today, the world is talking about agentic AI, and we are already saying that unless you cure the LLM or the brain of the system and make it more context-relevant, agents will not work,\u201d said Malhotra.<\/p>\n

India should work on world models \u2014 that understand cause-and-effect relationships in a given situation \u2014 using diverse data specific to the country. The next two-three years are very important in this regard as players will combine the idea of contextual (reasoning) models and world models, and then see how quantum computing joins this space, said Malhotra.<\/p>\n

He pointed out that India still lacks in terms of R&D for AI, a situation worsened by the lack of talent sovereignty and capability sovereignty in the country. <\/p>\n

\u201cIf India has to lead the race for AI, talent has to be created and utilised within the country. There has to be an ecosystem of investment in skills, education and applied research. If you notice, the global landscape of sovereignty, everyone has it. Europe has a regulation-led trust, which is the European Act, and they follow their own structure,\u201d said Malhotra.<\/p>\n

Great opportunity <\/h2>\n

Stating that there is great opportunity in AI world models, Sanchit Vir Gogia, Chief Analyst at Greyhound Research, voiced confidence in India\u2019s AI roadmap, dismissing the recent fall in IT stocks as \u201ca knee jerk reaction\u201d to global announcements including Anthropic\u2019s new AI plug-ins.<\/p>\n

In its latest report on technology reimagination, NITI Aayog called talent, business environment and R&D \u201ccritical enablers\u201d for scaling India\u2019s software and SaaS ecosystem.<\/p>\n

\u201cThe current technology talent pipeline remains service-oriented, requiring a shift toward product lifecycle capabilities and skills in areas like DevSecOps and product management. The ease of doing business also needs improvement, as complex regulations continue to constrain SaaS operations. Streamlined procedures for IP registration, export compliance and talent retention would enhance competitiveness,\u201d said the report.<\/p>\n