Project Tapestry has reached a key technical milestone while expanding its sovereign AI collaborations with Vietnam and India during United Nations General Assembly Week in New York. The initiative is developing an open, global consortium aimed at enabling nations and institutions to collaborate on advanced AI while retaining control over their data, sovereign models and deployments.
First Technical Milestone Demonstrates Consortium Training
Project Tapestry recently completed its first major technical milestone, demonstrating building blocks that allow organisations across different countries, infrastructure and jurisdictions to collaboratively train and adapt AI models.
Two proofs of concept coordinated model training across four geographically distributed sites, while the underlying training data remained local to the participating sites. The teams also demonstrated cultural alignment by modifying a large language model to better understand the unique social and linguistic aspects of Indian and Vietnamese cultures.
AI Training Without Centralising Data
The milestone provides an initial demonstration of one of Project Tapestry’s central premises: organisations do not need to centralise their computing resources or data under a single institution to collaboratively build increasingly capable AI models.
The approach allows participating organisations to work together while keeping their training data within their respective locations.
Vietnam Advances Sovereign AI Collaboration
During UN General Assembly Week, AI Alliance and Project Tapestry leaders met in New York with Vietnam’s General Secretary and senior government ministers to advance collaboration on developing sovereign AI capabilities for Vietnam.
The discussions affirmed an intention to collaborate on a sovereign AI model aligned with Vietnam’s national AI strategy and priorities. The planned focus includes Vietnamese language, culture, knowledge and national use cases.
Vietnam has participated in Project Tapestry since its early stages. The expanded collaboration marks a step toward applying the consortium model to national-scale sovereign AI development.
India Expands Its Role in Project Tapestry
During UN General Assembly Week, Consul General Binaya Srikanta Pradhan hosted Project Tapestry leaders at the Consulate General of India in New York.
Prof. Ganesh Ramakrishnan of BharatGen, Prof. Ritwik Banerjee of Stony Brook University in the United States, and Kaushik Bhatta, Partner and Director, AI Alliance, briefed the Consul General on consortium-driven AI development, Project Tapestry’s approach and Stony Brook University’s recent addition to the consortium.
India’s Languages and Cultures Highlighted
“India is a microcosm of the world. AI built for its languages and cultures preserves agency,” said Pradhan.
India has been an early contributor to Project Tapestry. BharatGen, India’s government-backed, open-source foundation model stack, conducted one of the two proofs of concept under the guidance of Dr Maneesh Singh, VP, Machine Learning at BharatGen, as part of Tapestry’s first technical milestone.
The proof of concept involved joint training with Monash University across India and Australia, while each institution’s data remained in its home country.
BharatGen Leadership Recognised at Goalkeepers New York
During the same week, the Gates Foundation named Prof. Ramakrishnan a 2026 Goalkeepers Champion, recognising him at Goalkeepers New York for his leadership of BharatGen.
The meeting also comes ahead of a Project Tapestry workshop in Mumbai on October 15, hosted by BharatGen. The workshop will bring together technical and institutional leaders to plan the next phases of India’s participation.
Project Tapestry Moves Towards Global Collaboration
With its first technical accomplishments and expanding national-scale collaborations, Project Tapestry is moving into its next phase. The initiative is expanding work around consortium training, cultural alignment and data governance, with growing participation from countries and institutions worldwide.
Tapestry Contributors Present Sovereign AI Research
Tapestry contributors also presented “Sovereignty Through Interdependence: Epistemic Agency, Intelligent Action, and Federated AI” at the First Workshop on Sovereign AI for Collaborative and Pluralistic AI Ecosystems.
The workshop was co-located with ACM HCOMP 2026 in Washington, D.C., and organised by researchers at Taiwan’s Academia Sinica.
The paper, authored by Dr. Maneesh Singh, VP, Machine Learning at BharatGen, Ritwik Banerjee, Lav R. Varshney and Kaushik Bhatta, examines how federated approaches such as Project Tapestry can support AI sovereignty through collaboration.
Leadership Comments
“Project Tapestry gives Viet Nam an opportunity to turn international collaboration into lasting national capability-developing AI that understands our language, culture, knowledge, and priorities, while building the engineers and institutions to operate, adapt, and advance the technology,” said Ambassador Hoang Anh Tuan, Consul General of Viet Nam in San Francisco.
“Our meeting with General Secretary and President Tô Lâm, and his strong support for this direction, underscores the broader vision for Tapestry,” said Christopher Nguyen, Chief Architect of Project Tapestry. “Across Viet Nam, India, Japan, Bhutan, and beyond, we are building a model for nations to shape the frontier together while preserving sovereignty over their data and technological future.”
“Our first milestone and growing global collaborations demonstrate that Tapestry is moving from architecture to working technology. Organizations operating across countries and infrastructure have now trained models together while maintaining control of their own data. It is an important first step toward building advanced AI collaboratively, openly, and sovereign by design”, said Anthony J. Annunziata, Director of Open Source AI at IBM.
Also read – UP Government to Support 50 AI Startups in Lucknow as HCL Tech Launches IT City Phase 2
Join us on LinkedIn for quick updates – FYI9 News LinkedIn
