Anthropic opens Singapore office and signs government AI agreement

Anthropic has opened a Singapore office, its fifth in the Asia-Pacific region, and signed a Memorandum of Understanding with Singapore’s Ministry of Digital Development and Information. The office will support customers, policy work, and hiring across the Association of Southeast Asian Nations. The agreement covers AI safety testing, evaluation standards, incident response, skills development, and adoption by government agencies and businesses.

Anthropic co-founder Ben Mann and ministry Permanent Secretary Kai Fong Chng signed the agreement during Singapore International Cyber Week. Dale Finlay, Anthropic’s general manager for ASEAN, leads the local team. Singapore Minister Josephine Teo and Anthropic APAC policy head Michael Sellitto witnessed the signing.

Demand arrived before the office

Anthropic’s Economic Index ranks Singapore second among 121 countries for Claude.ai usage relative to population. Usage is nearly six times the level its population alone would predict. Anthropic also cites several large deployments:

• More than 1,500 engineers and technical staff at GovTech Singapore use Claude Code, Anthropic’s coding assistant, to modernize government systems and accelerate software delivery.
• DBS uses Claude for software development and productivity workflows.
• GIC is rolling out Claude across its investment and corporate services teams.
• Grab says BriX, its Claude-powered sales assistant, saves sales teams about 40,000 hours each quarter.
• Sea uses Claude to improve Shopee’s customer-service chatbot and help thousands of employees write and debug code.

The pact targets testing and incident response

The agreement calls for technical exchanges between Anthropic and Singapore’s AI safety, security, and assurance institutions. Planned work includes model evaluations, testing standards, and incident-response practices. Evaluations measure a model against defined safety and performance criteria, while red teaming asks testers to probe for harmful, unreliable, or exploitable behavior.

The organizations began working together in 2023, when Anthropic joined the Infocomm Media Development Authority’s AI evaluation sandbox. They later conducted a multilingual red-team project covering English, Mandarin, Malay, and Tamil.

The agreement also commits Anthropic to support AI training, help small and medium-sized businesses adopt the technology, and share Economic Index data with Singapore as the government develops AI policy.

A local route for support and deployment

• Product surface: Anthropic announced no changes to Claude’s APIs, model lineup, or pricing.
• Regional support: ASEAN companies now have a local team for enterprise deployment, account support, and reliability feedback.
• Data processing: Organizations can use in-region Claude inference through Amazon Bedrock, Amazon’s managed AI platform. Processing within an AWS region can help regulated companies satisfy data-handling and residency requirements.
• Evaluation practices: The planned technical exchanges could influence how regional institutions test and approve AI systems for regulated workloads.
• Hiring and community: Anthropic is recruiting for engineering, policy, and go-to-market roles in Singapore. It has also hosted a Builder Summit for more than 300 developers and a community meetup attended by more than 700 local builders.