Michael Heinrich, CEO and Co-founder of 0G
For most artificial intelligence companies, Singapore is either a regional sales base or a policy-friendly testbed. For Zero Gravity, better known as 0G, the city-state is now becoming something more central: its global headquarters.
The AI infrastructure company said it will invest about US$15.4 million in Singapore over the next five years, using the country as a base to grow its operations, research partnerships and developer ecosystem.
The announcement was made at 0G Dev Day Singapore 2026 at the National Gallery Singapore, where close to 3,000 participants registered, including developers, researchers, businesses, investors and members of the wider technology community.
Also Read: SEA’s venture capital shifts from mega-rounds to AI and SaaS
The move places 0G in a country that has been working to position itself as a trusted AI hub for Southeast Asia and beyond. Singapore’s National AI Strategy 2.0 has put emphasis on talent, compute access, governance and practical adoption, while the city’s regulators have tried to balance innovation with safety. For companies building AI infrastructure, that mix is attractive: Singapore offers proximity to regional markets, deep pools of capital and a government keen to shape global AI standards.
Senior Minister of State for Digital Development and Information and Health Tan Kiat How, who delivered the keynote at the event, framed Singapore as a place where AI products can be deployed, adapted and scaled. That message fits neatly with 0G’s own pitch: build in Singapore, then export globally.
Michael Heinrich, CEO and co-founder of 0G, said the company’s commitment is “not simply about opening an office” but about “research, talent, innovation and long-term investment”. He added: “We see Singapore not as a final destination, but really as a launchpad for the rest of the world.”
A headquarters play, not just an office opening
Of the planned US$15.4 million investment, around US$7.7 million will go towards establishing Singapore as 0G’s global headquarters, building operations, growing the team and developing its ecosystem.
The company has already committed about US$3.9 million to a four-year research collaboration with Nanyang Technological University (NTU), focused on decentralised AI and blockchain-based infrastructure. It plans to commit another US$3.1 million to research collaborations with other universities and academic partners. A further US$770,000 will support AI and badminton research with NTU and YB Badminton Academy.
0G’s first physical office globally is also in Singapore, at Stamford Place. The location is symbolically convenient: the building formerly housed Singapore’s National Heritage Board, a detail Heinrich has used to position the company as one building future infrastructure from a site linked to the country’s past.
Also Read: Singapore’s Psalion raises US$50M fund for Web3’s next practical phase
At a practical level, the decision reflects a broader pattern in Southeast Asia’s tech ecosystem. Singapore continues to attract regional headquarters for startups, fintech firms, AI companies and crypto-related infrastructure players, even when their markets, users and developer communities are spread across Indonesia, Vietnam, the Philippines, India and beyond. Its strengths lie less in domestic market size and more in regulation, talent mobility, institutional partnerships and investor access.
Taking decentralised AI from lab to court
The more unusual part of 0G’s announcement is not the headquarters plan, but where some of its research is going next: badminton.
At Dev Day, Zero Gravity Labs and NTU formalised the next phase of their existing US$3.9 million research collaboration, expanding work in decentralised AI towards real-world applications. Sports will be an initial focus, with badminton as the starting point.
The partnership, established in 2025, has been working on decentralised AI training, model alignment and verifiable AI systems. Researchers from 0G and NTU have had three collaborative research papers accepted at NeurIPS and one at ICML, two of the most closely watched machine learning conferences globally.
Badminton is not a random choice. In Southeast Asia, the sport has deep cultural and competitive significance, particularly in Indonesia, Malaysia, Thailand and Singapore. Singapore’s Loh Kean Yew, a former world champion, has also helped raise the sport’s profile locally, while his Asian Games silver added further momentum.
For AI researchers, badminton offers dense and fast-moving data. A single rally can involve rapid changes in body position, shuttle trajectory, footwork, shot selection, fatigue and tactical decision-making. Much of that is hard to capture in one clean dataset.
Working with NTU and YB Badminton Academy, 0G plans to explore a badminton foundation model that combines data from players, movement, video, training and competition. Foundation models are AI systems trained on broad datasets and then adapted for specific tasks. In this case, possible uses include helping coaches spot movement patterns that are difficult to see in real time, identifying training loads that may increase injury risk, and supporting young players with more personalised feedback.
“Badminton brings together frontier AI research, real-world data and a very human outcome: helping people perform better and stay healthier over time,” Heinrich said.
Also Read: Malaysia’s sovereign AI bet: Local context becomes the next startup moat
0G Singapore also signed a three-year memorandum of understanding with YB Events and Sports for Project New Frontiers: Bringing Badminton to New Horizons. Under the partnership, 0G Singapore will commit about US$204,000 annually for three years to support youth badminton development, international competition and the exploration of AI and emerging technologies in sport.
The trust layer question
Beyond sport, 0G is pitching itself as part of the infrastructure stack needed for safer AI adoption. The company describes its mission as building the “trust layer for AI”, with decentralised infrastructure designed to make AI more private, verifiable, sovereign and open.
At Dev Day, it showcased 0G Private Computer, a product aimed at bringing privacy and verification to sensitive AI workloads. Heinrich said the system is designed so that sensitive workloads can remain private, including from 0G itself. The broader idea is to give organisations more confidence over how data is handled and how AI workloads are processed, especially when proprietary information is involved.
The company also introduced BestCEO, a ready-to-use AI solution aimed at making advanced AI capabilities accessible to businesses, including small and medium-sized enterprises, without requiring them to build AI infrastructure or applications from scratch.
This is a relevant angle for Southeast Asia, where SMEs make up the backbone of most economies but often lack the technical teams or budgets to adopt AI meaningfully. The region’s AI opportunity is not only about large enterprises deploying copilots, but also about whether smaller companies can use the technology without handing over sensitive data or depending entirely on foreign platforms.
Rivals in a crowded AI infrastructure race
0G is entering a crowded and fast-shifting market. In decentralised AI and compute, it sits alongside global projects such as Bittensor, Gensyn, Akash Network, Render Network and Aethir, all of which are trying to rethink how AI models, compute resources and data networks are coordinated. It also competes indirectly with hyperscale cloud providers such as Amazon Web Services, Microsoft Azure and Google Cloud, which remain the default infrastructure providers for most AI companies. The key question for 0G is whether developers and enterprises will see decentralisation and verifiability as must-have infrastructure, rather than a technically elegant alternative to existing cloud systems.
Community as a test of accessibility
0G is also trying to position AI access as a community issue, not just an enterprise problem. At its first community AI workshop in Singapore, 70 residents, including many seniors, learned about AI and built websites and applications on 0G over two days. One senior participant built a website to help people understand which health screenings they should consider at different ages, including reminders for women about mammograms.
Also Read: Crypto is dead? Apparently not, says Y Combinator – Blockchain is still worth building
That example is small, but it points to the broader challenge around AI adoption in the region. For AI to move beyond boardrooms and developer conferences, it must become useful to ordinary users, schools, coaches, seniors and small businesses.
As 0G plants its headquarters in Singapore, the company is making a familiar but ambitious bet: that the city-state can serve as a neutral, trusted base for technologies meant to travel far beyond its borders. The harder test will be whether its infrastructure can move as smoothly from research papers to badminton courts, SMEs and real-world AI systems.
The post 0G picks Singapore as global base, expands AI research into badminton appeared first on e27.