Anthropic opened an office in Singapore on October 14, 2026, its fifth base in the Asia-Pacific region after South Korea, Japan, India and Australia, according to The Straits Times. Alongside the launch, the developer of Claude agreed a memorandum of understanding (MOU) with Singapore's Ministry of Digital Development and Information (MDDI). CNA reported that co-founder Ben Mann and MDDI Permanent Secretary Chng Kai Fong signed it a day earlier at Singapore International Cyber Week, with Minister Josephine Teo and Anthropic's Asia-Pacific policy head Michael Sellitto as witnesses.
The document groups the cooperation into three strands: safety, security and model evaluation; research on how AI is being taken up and what it changes; and responsible use across government and business. Anthropic says the first strand will take the form of technical exchanges with Singapore's AI safety, security and assurance bodies on evaluation methods, testing standards and incident response. The company also intends to share research and data with officials, including figures from its Anthropic Economic Index, and to advise as Singapore writes its own AI rules.
Disclosure is the sensitive point. CNA noted that Senior Minister K. Shanmugam had said the day before that Singapore wants immediate transparency from AI firms when something goes wrong. Sellitto told reporters that Anthropic would report cyber incidents as fast as it responsibly can. He added that Singapore receives no early access to unreleased Anthropic models, though local organisations may apply to the company's cyber verification programme to test models that have already shipped.
Commercial demand explains much of the timing. In its latest Economic Index, Anthropic ranks Singapore second of 121 countries for Claude.ai use relative to population, at almost six times the level its size would suggest. The company says more than 1,500 GovTech Singapore engineers and technical staff use Claude Code on government systems, that DBS applies Claude to software work and daily productivity, and that the sovereign fund GIC is deploying it across investment and corporate teams. Grab, Sea and Supabase are also cited. These adoption numbers are Anthropic's own and were not independently checked.
Anthropic would not give a headcount for the Raffles Place office. The Straits Times and CNA each counted 12 local vacancies on its careers site, spanning sales, marketing, recruitment and finance, and ASEAN general manager Dale Finlay told CNA the near-term priority is go-to-market work. GIC is also an investor: The Straits Times reports it joined a 2025 round and led a $30 billion Series G in February 2026.
Singapore has been collecting agreements of this kind. CNA points to a deal with OpenAI in May, an AI safety MOU signed in June by Microsoft and the Infocomm Media Development Authority (IMDA), and the scam-detection model GovTech is building with Google DeepMind. Anthropic's own ties go back to 2023, when it entered the Generative AI Evaluation Sandbox operated by IMDA together with the AI Verify Foundation, followed by red-teaming in English, Mandarin, Malay and Tamil.
Neither party has released the MOU text, a budget, timelines or measurable deliverables. Whether the testing exchanges and incident-sharing promises lead to published results, joint standards or faster disclosure is the part observers will be able to judge only over the coming months.