Anthropic opens a Singapore office — and signs an AI safety pact with the government
Anthropic has opened a Singapore office — its fifth in Asia-Pacific — and signed a memorandum of understanding with Singapore’s Ministry of Digital Development and Information (MDDI) covering AI safety, security, model evaluation, and responsible adoption. The company announced both moves on October 14, 2026, calling Singapore one of its most active markets anywhere.
- The MOU was signed by Kai Fong Chng, Permanent Secretary of MDDI, and Anthropic co-founder Ben Mann on the sidelines of Singapore International Cyber Week; Minister Josephine Teo and Anthropic's Michael Sellitto witnessed.
- It sets out cooperation on AI safety, model evaluation, testing standards, and incident response, plus skills development, responsible SMB adoption, AI capability among public officers, and research sharing — including from the Anthropic Economic Index.
- Anthropic has worked with Singapore since 2023 (the Generative AI Evaluation Sandbox run by IMDA and the AI Verify Foundation, later a red-teaming project spanning English, Mandarin, Malay and Tamil); Singapore also signed an agreement with OpenAI in May 2026 and an IMDA–Microsoft safety memorandum in June.
- The Singapore office is its fifth in Asia-Pacific, joining Tokyo, Bengaluru, Seoul and Sydney; GM ASEAN Dale Finlay said sales and market development are the immediate priority — no core research operation planned yet. The office is open and hiring, with at least a dozen roles advertised October 14.
- Anthropic's Economic Index ranks Singapore second among 121 countries for Claude.ai usage relative to population; 1,500+ GovTech engineers use Claude Code; the Cyber Security Agency is exploring Claude for national cyber defense.
- Private sector: DBS runs Claude across development and productivity; GIC is rolling it out; Grab's Claude-powered BriX sales assistant saves ~40,000 hours per quarter; Sea uses Claude for customer-service chatbots and has thousands of employees writing and debugging code with it.
- Michael Sellitto stressed the agreement provides for testing but gives Singapore no preferential access before models are commercially released.
The memorandum was signed by Kai Fong Chng, Permanent Secretary of MDDI, and Anthropic co-founder Ben Mann on the sidelines of Singapore International Cyber Week. Minister for Digital Development and Information Josephine Teo and Michael Sellitto, Anthropic's head of APAC policy, witnessed the signing. The agreement sets out cooperation on AI safety, model evaluation, testing standards, and incident response, plus research on how AI adoption plays out across the public and private sectors.
Signed on the sidelines of Cyber Week
Under the pact, Anthropic will hold technical exchanges with Singapore's AI safety, security, and assurance institutions on model evaluation, testing standards, and incident response — work the company says mirrors what it already does with safety institutes in other countries. The memorandum also covers skills development, helping small and medium businesses adopt AI responsibly, building AI capability among public officers, and sharing research and data — including from the Anthropic Economic Index — as Singapore writes its AI rules. "We founded Anthropic on the belief that capability and safety have to go hand in hand," Mann said in the announcement.

The relationship isn't new. Anthropic has worked with Singapore since 2023, when it joined the Generative AI Evaluation Sandbox run by the Infocomm Media Development Authority and the AI Verify Foundation, later partnering with both on a red-teaming project spanning English, Mandarin, Malay, and Tamil. Singapore has been collecting frontier-lab partnerships at a steady clip: an agreement with OpenAI in May 2026 and a safety-and-security memorandum between IMDA and Microsoft in June.
Singapore punches far above its weight on Claude
The commercial case is unusually concrete. Anthropic's Economic Index ranks Singapore second among 121 countries for Claude.ai usage relative to population — nearly six times what its size alone would predict. More than 1,500 engineers and technical staff at the Government Technology Agency of Singapore (GovTech) use Claude Code to modernize government systems, and the Cyber Security Agency of Singapore is exploring how Claude could strengthen national cyber defense.
The private-sector list is just as dense. DBS — Southeast Asia's largest bank — runs Claude across software development and productivity inside its security and governance framework. GIC, Singapore's sovereign wealth fund, is rolling Claude out across investment and corporate services teams; organizations in regulated sectors can now run it with in-region inference on Amazon Bedrock. Grab's Claude-powered sales assistant BriX saves its sales teams roughly 40,000 hours per quarter, while Sea — the company behind Shopee, Garena, and Monee — uses Claude for a faster customer-service chatbot and has thousands of employees writing and debugging code with it. Developer platform Supabase, rooted in Singapore, is building with Claude Code.

What each side gets — and what it doesn't
For Anthropic, Singapore is an office, not a research lab — at least for now. General manager of ASEAN Dale Finlay said sales and market-development capabilities are the immediate priority; managing director of international Chris Ciauri said a core Singapore research operation is not currently planned, though that could change after an assessment of local talent. The office is open and hiring, with at least a dozen Singapore roles advertised on October 14, and Anthropic marked the launch with its Builder Summit Singapore, drawing more than 300 developers. Mann said efforts are underway to bring the company's frontier AI models to Singapore without giving dates; Sellitto stressed the agreement provides for testing but gives Singapore no preferential access before models are commercially released.
For Singapore, the play is influence before access: shared technical information that gives its safety institutions a clearer view of frontier capabilities and risks, while developers get safer models. Whether the testing-first model survives the jump from memorandum to practice will be the number to watch in the next year.
Frequently asked questions
What did Anthropic and Singapore agree to?
A memorandum of understanding between Anthropic and Singapore's Ministry of Digital Development and Information covering AI safety, model evaluation, testing standards, incident response, skills development, responsible SMB adoption, and research sharing — signed on the sidelines of Singapore International Cyber Week.
Is Anthropic opening a research lab in Singapore?
No — the new office, its fifth in Asia-Pacific, is a sales and market-development operation for now; GM ASEAN Dale Finlay said a core Singapore research operation is not currently planned, though it could change after a talent assessment.
Why Singapore?
Anthropic calls it one of its most active markets anywhere: Singapore ranks second of 121 countries for Claude.ai usage relative to population, 1,500+ GovTech engineers use Claude Code, and DBS, GIC, Grab and Sea are all heavy Claude users.