中文标题待翻译

Google AI Futures Fund co-founder and Director Jonathan Silber (L) with Granite Asia’s Senior Managing Partner Jenny Lee

Asia’s AI founders are increasingly facing a split-screen reality. Investor interest is high, but building at the frontier of artificial intelligence still requires more than a seed cheque. Startups need access to large models, cloud infrastructure, technical feedback, distribution channels and patient capital, all while competing with better-funded peers in the US and China.

Granite Asia and Google AI Futures Fund are trying to package some of those ingredients into a single programme.

Also Read: PSA, Granite Asia launch US$50M fund to scale supply-chain innovation

The Singapore-headquartered investment platform has announced a strategic partnership with Google AI Futures Fund to co-develop and co-invest in AI startups across Asia. The two sides plan to invest up to US$2 million per company, with founders also receiving access to Google’s AI models, cloud credits and technical teams.

The programme will be routed through the Granite Asia Fellows Program, the firm’s selective platform for founders building global companies from Asia. Startups will be sourced by Granite Asia and evaluated jointly by both firms, though investments will be made through their own independent processes and on identical economic terms.

Granite Asia says it has around US$11 billion in assets under management and co-managed capital. The firm, which has backed technology companies across Asia for more than two decades, is positioning the tie-up as a way to connect regional founders with global AI infrastructure at an early stage.

Capital, models and technical access

In addition to US$2 million in co-investment, selected startups will also receive up to US$350,000 in Gemini and Google Cloud credits. They will also get early access to Google’s AI models, including Gemini, Nano Banana and Lyria, as well as support from product, research and engineering teams across Google, including Google DeepMind.

This is crucial because AI startups face unusually high upfront costs. Unlike many software companies of the previous decade, AI-native businesses often need expensive computing power, access to high-performing models, domain-specific data and engineering support before they can prove commercial traction. For early-stage founders in Southeast Asia and the wider region, these constraints can be particularly sharp.

Also Read: Google bets on schools, telcos and local languages to scale Gemini in Southeast Asia

Cloud credits are not the same as revenue, but they can extend the runway for young companies experimenting with model training, inference or product deployment. Access to technical teams may also help founders avoid months of trial and error as they decide whether to build on top of existing large language models, fine-tune models for specialised use cases, or combine several systems into agentic workflows, software agents that can plan and execute tasks with limited human input.

The partnership will focus on five broad areas: AI for science and knowledge, AI for industrial transformation, AI for work, AI for creativity, and AI for entertainment. These are wide categories, but they point to sectors where Asian markets may offer both demand and data: manufacturing, logistics, education, healthcare, media, gaming, enterprise services and consumer platforms.

Jonathan Silber, co-founder and Director of Google AI Futures Fund, said Asia’s industrial depth, scientific talent and consumer scale make it “a distinct frontier for what AI can build”. He added that founders in the programme would gain “direct collaboration with Google DeepMind researchers and engineers” as well as early access to advanced models.

Jenny Lee, Senior Managing Partner at Granite Asia, framed the partnership as a bridge between global AI capability and Asian founder ambition. “The next generation of category-defining AI companies will be built in Asia — grounded in the region’s industrial depth, scientific talent, and consumer scale, then scaled globally,” she said.

Why Asia is becoming a serious AI testbed

The announcement comes as Southeast Asia’s AI ecosystem moves from experimentation to deployment. Banks are using AI for compliance and customer service, manufacturers are exploring predictive maintenance and quality inspection, and consumer internet companies are adding generative tools to content, commerce and gaming products.

The region also has characteristics that make AI adoption complicated but potentially valuable. Southeast Asia is multilingual, mobile-first and highly fragmented by regulation, culture and market maturity. A product that works in Singapore may need significant localisation for Indonesia, Vietnam, Thailand or the Philippines. This gives regional founders a chance to build AI products that understand local workflows, languages and constraints rather than simply importing tools designed for Western enterprise customers.

At the same time, the gap between ambition and capability remains wide. Many startups lack access to frontier models, high-quality training data, senior AI engineering talent and customers willing to pay for experimental products. The challenge is not just building demos, but turning them into reliable products that enterprises or consumers use repeatedly.

This is where platform partnerships have become more important. For global AI companies, working with investors gives them a route to promising founders before those startups become large customers or strategic partners. For venture firms, access to model providers can make their capital more attractive in a crowded funding market. For founders, the best-case outcome is faster product development without being locked into a single technical path.

Also Read: Ecosystem Roundup: The AI funding boom has a Southeast Asia-sized blind spot

Granite Asia says founders in the programme will retain full technical choice across their AI stack. That detail is important. AI founders are increasingly wary of building too deeply around one provider’s infrastructure, especially as model performance, pricing and open-source alternatives shift quickly.

A crowded race for AI founders

The partnership lands in a competitive investment landscape. Across Asia, venture firms such as Peak XV Partners, Lightspeed, Accel, Antler, East Ventures, Wavemaker Partners, Vertex Ventures, Insignia Ventures Partners and Openspace Ventures are all searching for AI companies that can move beyond thin application layers. Globally, firms including Andreessen Horowitz, Sequoia Capital and Index Ventures have also leaned hard into AI, often with larger pools of capital and closer proximity to Silicon Valley’s model labs.

Strategic capital is becoming just as important. Microsoft, Amazon Web Services, Nvidia and Google are all courting AI startups through credits, technical support, model access and go-to-market programmes. The question for Asian founders is whether these partnerships translate into durable advantages or simply subsidise early infrastructure costs before startups must compete on product, data and distribution.

Granite Asia’s bet appears to be that Asia can produce more than localised AI applications. The firm is looking for companies that can start with regional advantages, such as industrial networks, scientific talent, consumer scale or multilingual markets, and then expand globally.

That is an ambitious bar. Many AI startups today risk being features rather than companies, especially when large model providers can quickly replicate popular use cases. The strongest candidates will likely be those with proprietary data, deep domain knowledge, strong customer relationships, or workflows that are difficult for general-purpose models to replace.

Also Read: Malaysia’s sovereign AI bet: Local context becomes the next startup moat

For Southeast Asia, the programme also reflects a broader shift in how the region is viewed. The old story was that major technology platforms were built elsewhere and adapted locally. The new pitch is that Asian founders can use the region’s complexity as a proving ground, building products that are resilient enough for fragmented markets and scalable enough for the world.

Whether this partnership produces such companies will depend less on the size of the cheque than on founder selection and follow-through. US$2 million, cloud credits and model access can give startups a strong start. But in AI, as in every technology cycle before it, the harder test comes later: finding real customers, building defensible products, and proving that the company is more than a clever demo.

The post Granite Asia, Google AI Futures Fund team up to back Asia’s AI founders appeared first on e27.