Singapore courts OpenAI, Google, and Anthropic, but Washington and Beijing complicate “neutral”

Singapore’s AI momentum has hit a point where “neutral hub” sounds less like marketing and more like a live strategic bet. In the past year, OpenAI and Google DeepMind established applied AI labs in Singapore, while Anthropic advertised local roles spanning finance, product support, and economic research. Meanwhile, Chinese firms have also stepped up: Tencent has deepened investment, and Singapore keeps showing up as the practical place where teams, customers, and partners intersect.

Fortune captures why in one blunt line from Bright Data’s chief revenue officer, Gunja Gargeshwari: it’s easiest to operate in the region if you have people in Singapore, because “it’s where conversations are happening.” That is the promise the city-state sells, and it’s why U.S. and Chinese AI companies are setting up shop there in the first place. The question executives now have to answer is whether that conversation advantage survives the latest geopolitical friction, especially when regulators start looking through holding structures and blocking model usage.

Start with the on-the-ground expansion. Bright Data, for example, positions Singapore as its APAC headquarters, even though 60% of its Asian customer base hails from China and India. Plaud, a San Francisco-based AI notetaker company, hired its first Singapore-based employee in 2024. On June 10, Plaud said it would spend 10 million Singapore dollars ($7.8 million) to expand local operations, and it plans to grow headcount from 100 to 150 by the end of the year. This is classic APAC playbook logic: be where the demos happen, where customer engineering and procurement cycles move faster, and where partnerships can be staffed locally.

Singapore’s pitch to the AI industry is not only operational convenience. It’s geopolitics, packaged as governance. The country markets itself as an “economic safe haven,” citing regulatory clarity and strong governance, and it leans on its long track record of being stable and predictable. Singapore Prime Minister Lawrence Wong framed the tradeoff in a policy conference last July, saying that even if the city-state is “boring” compared with New York and Paris, it is “stable,... predictable.... reliable and... trusted,” and those intangible assets are “intangible assets that others would die to have.” For AI founders, that reliability matters because the AI market is moving from building models to monetizing them. As BNY’s wealth analysts wrote in a March report, the defining feature of the AI cycle through 2025 was capital expenditure that expanded capacity but invited skepticism, with attention shifting “decisively from scale to return on investment.” Local presence is increasingly tied to revenue, not just research.

Now add the labor and talent ecosystem. Founders point to Singapore’s education pipeline and the difficulty of hiring top engineers. Plaud CEO Nathan Xu says the biggest pain for him and the company is hiring the best engineers, and that Singapore is home to some of the best universities in the world. In this year’s QS World University Rankings, the National University of Singapore ranked #8, while the country’s Nanyang Technological University came in at #12. Xu also describes a pipeline that curates generations of talent around software engineering, computer science, AI, data science, and operations.

For U.S. AI firms, Singapore is also a go-to-market bridge. OpenAI opened a regional office in Singapore in 2024. Last month, it committed 300 million Singapore dollars ($234 million) to growing the country’s AI ecosystem. It also announced the opening of an applied AI lab, the first outside of the U.S., meant to make Singapore one of its hubs for forward deployed engineers. These are specialized software engineers who embed directly within customer organizations to customize and deploy solutions. Notion, the AI-powered productivity platform, opened a Singapore office in mid-2025, and Randy Hunt, its head of design, said the priority is to meet and interface with current and potential customers. He adds that a video demo can work, but doing it next to a customer “resonates better.” Enterprise AI makes Singapore even more attractive because many multinational companies house APAC headquarters there.

But the story has an uncomfortable second act: neutrality is getting stress-tested. Manus AI and its parent company, Butterfly Effect, relocated their global headquarters to Singapore in mid-2025 to avoid Western regulatory scrutiny and to better access global capital. In December, Manus AI sold itself to Meta for $2 billion. Beijing quickly moved to block the deal, and in April ordered the acquisition to be unwound. The Singapore holding structure did not save the transaction. Manus’s legal status as a Singapore company didn’t matter because its continued footprint in China gave Beijing jurisdiction. Sebastian Wiendieck, head of legal practice in China at law firm ROEDL, told CNA that regulators “looked straight through the Singapore holding structure to the technology’s Chinese origin,” calling it “a new normal” in which any China-founded AI startup, regardless of offshore domicile, faces “intense national security scrutiny” if it tries to sell to a U.S. buyer.

The U.S. can also directly limit what Singapore-based teams can access. Last week, the U.S. government barred non-U.S. individuals from using Anthropic’s powerful Mythos model. That creates a risk that Singapore could lose access to powerful frontier models from U.S. companies like Anthropic and OpenAI, depending on how the restrictions apply. Even with these cracks, companies remain positive about expanding into Singapore. The country released its national AI R&D plan in January, alongside a 1 billion Singapore dollar injection to fund AI-related infrastructure and capabilities. It also set plans to build an AI industrial park called Kampong AI, opening in 2028 with workspaces and housing facilities to attract AI start-ups. Xu says the feeling is “we are welcomed here,” and he highlights the scale of the ramp-up, from “zero people here” a year ago to “close to a hundred.”

For executives making similar decisions, the practical takeaway is that Singapore is still a conversation hub, but now conversations come with compliance gravity. The upside is obvious: faster customer access, talent density, and a regulatory posture that many firms can operationalize. The downside is equally real: regulators in both Washington and Beijing can override corporate structure and restrict model usage, turning “neutral” into a moving target. The best move for AI leaders is not to stop betting on Singapore, but to understand that the board-level question has changed. It is no longer just “Can we scale here?” It is “Can we scale here through the regulatory lenses that follow our origin, our customers, and our technology?”