Anthropic Opens Singapore Office, Partners With Government on AI Safety
The company's fifth Asia-Pacific office comes with a memorandum of understanding with Singapore's government. Anthropic will help public agencies and businesses use Claude, while the two sides work together on AI safety evaluations and workforce training.
Anthropic announced on October 14 that it has opened an office in Singapore and signed a memorandum of understanding with the Singapore government. It is the company’s fifth office in the Asia-Pacific region. The agreement is with Singapore’s Ministry of Digital Development and Information. The two sides will work together on AI safety and security, model evaluations, and research into the societal effects of AI adoption.
Anthropic said it will build out its local team to help businesses and public sector agencies put Claude to work. Dale Finlay, who leads Anthropic’s ASEAN business, said customers in Singapore demand a high bar for reliability and security. Having a team on the ground, he said, lets the company tackle customers’ toughest problems up close. The office is already operating, and hiring is underway.
What the government partnership covers
The agreement expands on earlier AI safety evaluation work. In 2023, Anthropic took part in a generative AI evaluation project run by Singapore’s Infocomm Media Development Authority and the AI Verify Foundation. The two sides later ran joint exercises probing for AI weaknesses in English, Mandarin Chinese, Malay, and Tamil.
Going forward, Anthropic will exchange technical know-how with Singapore’s AI safety and security agencies. Methods for evaluating models, testing standards, and incident response are the main items on the agenda. Anthropic said it will share research and data with the government, including its own economic index. The company also said it will contribute technical and policy expertise as Singapore develops rules around AI.
The agreement also includes work aimed at the people and organizations actually using AI. Anthropic plans to support small and medium-sized businesses in adopting AI responsibly, and to explore ways to strengthen AI skills among civil servants.
Where Claude is being used
GovTech Singapore, the government’s technology agency, uses Claude Code for development and system maintenance work. According to Anthropic, more than 1,500 engineers and technical staff have adopted it. The company said the tool has sped up both improvements to government systems and software development. The Cyber Security Agency of Singapore is examining whether Claude can be applied to national cyber defense.
In financial services, DBS uses Claude for software development and day-to-day work across its Singapore operations. GIC, Singapore’s sovereign wealth fund, is rolling it out to its investment and corporate support teams. Anthropic said that financial institutions and other organizations can have Claude’s responses generated within the region when running on Amazon Bedrock. That means even organizations under strict data processing rules can meet their requirements, the company said.
How much time is being saved
Anthropic said BriX, a sales assistant tool built by Grab, uses Claude to save the sales team roughly 40,000 hours of work per quarter. BriX is also used to surface new sales opportunities. Sea, the company behind Shopee, uses Claude to improve the speed and accuracy of its customer service chatbot. Thousands of Sea employees also use it to write code and fix bugs.
In Anthropic’s most recent economic index, Singapore ranked second out of 121 countries for Claude.ai usage relative to population. The new office places support staff close to organizations that are already using Claude. What to watch next is how the evaluation-standards collaboration and AI skills initiatives in the agreement are actually carried out, and how companies measure their reported time savings in day-to-day work.