Table of Contents Show
As artificial intelligence continues to reshape industries and economies worldwide, OpenAI has made two significant announcements that underscore its long term commitment to India and responsible AI deployment globally. The company has appointed Prabhjeet Singh as Managing Director for India and simultaneously launched GPT-5.6, its most safety hardened model family to date, currently available in a limited preview.
A Strategic Leadership Move
Prabhjeet Singh brings an exceptional professional pedigree to his new role. A graduate of IIT Kharagpur and IIM Ahmedabad, Singh spent nearly a decade at McKinsey & Company as an Associate Partner, advising clients across financial services, telecom, and consumer technology, before joining Uber India in 2015 as General Manager and Head of Strategy. By 2020, he had risen to President of Uber India and South Asia, overseeing the platform’s mobility operations across India, Sri Lanka, and Bangladesh, while spearheading its transition into electric mobility and integration with India’s digital public infrastructure.
Singh will officially assume his responsibilities at OpenAI in September 2026, reporting directly to Kiran Mani, OpenAI’s Managing Director for the Asia Pacific region. His mandate spans consumer growth, enterprise adoption, strategic partnerships, regulatory engagement, and overall operations. With India recognised as OpenAI’s second largest market after the United States, this appointment reflects a clear and deliberate strategic commitment to the region.
Raising the Bar on AI Safety
Alongside the leadership announcement, OpenAI unveiled the GPT-5.6 model family, comprising three tiers — Sol, the flagship model; Terra, a balanced option for everyday enterprise work; and Luna, a fast and cost efficient variant. GPT-5.6 Sol, the most advanced of the three, introduces a new maximum reasoning mode and an ultra mode that deploys subagents for complex workflows. The entire family launches with what OpenAI describes as its most robust safety stack to date, backed by more than 700,000 A100-equivalent GPU hours of automated red teaming to uncover jailbreak techniques prior to release. Access currently remains limited to a select group of trusted partners, with broader availability expected in the coming weeks.
A Calculated Dual Push
Taken together, these announcements reflect a coherent strategy — investing in local leadership to accelerate market penetration while deploying technically rigorous safeguards that enterprise and institutional clients increasingly demand. OpenAI’s moves this week suggest a company thinking beyond growth, and toward enduring relevance.
Also Read: BharatGen to Anchor India’s Participation in Project Tapestry Global AI Consortium
