i
News
All Research News
News · 2026-10-14

Anthropic opens Singapore office and signs AI cooperation pact

@neuronium_ai @neuronium_ai

Anthropic has opened an office in Singapore and signed a memorandum of understanding with the government, combining a larger local presence with cooperation on AI safety and adoption. The agreement covers model testing, research into AI’s effects and responsible use in public and private sectors. It formalizes work that began in 2023, as Singapore emerges as one of Claude’s most intensive markets: its per-capita use ranks second among 121 countries.

Cover: Anthropic opens Singapore office and signs AI cooperation pact

A partnership with a track record

The memorandum was signed by Kai Phong Chng, permanent secretary at the Ministry of Digital Development and Information, and Anthropic co-founder Ben Mann, during Singapore International Cyber Week. Minister Josephine Teo and Michael Sellitto, Anthropic’s head of policy for Asia-Pacific, attended.

The agreement lays out cooperation on:

AI safety, model protection and evaluation
Research into AI adoption and its effects
Responsible use of AI in government and business

This is not a partnership starting from scratch. Anthropic joined a generative AI testing initiative organized by the Infocomm Media Development Authority and AI Verify Foundation in 2023. It later worked with them on testing models’ resilience to attacks in English, Chinese, Malay and Tamil.

Under the new memorandum, Anthropic plans technical consultations with Singaporean institutions responsible for AI safety and model evaluation. Discussions will cover evaluation methods, testing standards and incident response. The company also plans to share research and data, including material from its economic index, and offer technical and legal expertise as the government develops AI rules.

Other commitments focus on building local capacity: exploring ways to develop AI skills, helping small and medium-sized businesses adopt AI responsibly, and supporting the government’s goal of strengthening civil servants’ capabilities.

Singapore is already a major Claude market

Anthropic’s latest economic index places Singapore second among 121 countries for Claude.ai use relative to population. Its usage is nearly six times higher than expected for a country of its size.

That demand is visible across government and business:

More than 1,500 engineers and technical specialists at GovTech Singapore use Claude Code to update government systems and speed software releases. The Cyber Security Agency of Singapore is also exploring how Claude could strengthen national cyber defenses.
DBS, Southeast Asia’s largest bank, uses Claude in Singapore across software development and everyday work. Sovereign wealth fund GIC is gradually introducing it in investment and corporate teams.
For companies in regulated industries, Claude inference is available regionally through Amazon Bedrock, helping meet strict data-handling requirements.
Grab uses Claude in BriX, a sales assistant that saves employees about 40,000 hours per quarter and helps identify sales opportunities. Sea uses it to improve Shopee’s customer-support chatbot, while thousands of its employees use Claude to write and debug code. Supabase uses Claude Code in its work.

The unanswered question behind the expansion

Anthropic’s opening week also included a Builder Summit in Singapore, attended by more than 300 developers and engineering and product leaders. Earlier in the week, local Claude community ambassadors held a meetup for more than 700 developers. The office is open, and the company is hiring.

Dale Finlay, Anthropic’s managing director for ASEAN, said the company was “committed to Singapore for the long term.” Co-founder Ben Mann said Singapore’s investment in model testing and evaluation makes it a leader in responsible AI. The company’s stated ambition is to help Singapore, the Asia-Pacific region and the world transition successfully to powerful AI systems.

I think the more consequential test is whether this arrangement can make safety cooperation concrete: the announcement names consultations, standards and incident response, but not how their results will be measured. Singapore’s unusually high Claude use gives both sides a substantial setting in which to do that work; the office and memorandum make the relationship closer, not the outcome certain.

Source

Daily AI news

Every day we pick what actually matters in AI and explain it plainly — no hype, no filler. Subscribe if you want to follow where the industry is going.

Only what matters — every day

Follow on X