Anthropic Opens Singapore Office, Signs AI Safety MOU with Gov

Anthropic's new Singapore office aims to enhance local support for its customer base, which includes major financial institutions like DBS and GIC, public sector entities such as GovTech Singapore, and private companies like Grab and Sea. The company noted that regulated firms can access Claude through in-region inference on Amazon Bedrock, which it claims assists them in meeting stringent data processing requirements.

The MoU outlines cooperation on AI safety, security, model evaluation, research into AI adoption, and responsible deployment across both public and private sectors. Anthropic plans to engage with Singaporean AI safety institutions on technical exchanges covering model evaluation, testing standards, and incident response. This suggests a focus on developing shared best practices and technical frameworks.

Beyond safety, the agreement also addresses AI skills development, supporting adoption among small and medium-sized enterprises, and building AI capabilities within the public sector. Anthropic intends to share research, data from its Economic Index, and provide policy expertise to aid Singapore's AI regulatory development. This builds on its prior involvement in the 2023 Generative AI Evaluation Sandbox and multi-lingual AI red-teaming exercises.

The range of reported customer applications, from software development at DBS and GovTech to sales assistance at Grab and customer service at Shopee, indicates a broad commercial and governmental integration of Anthropic's AI. The stated high standards for reliability and security among Singaporean customers, as noted by Anthropic's General Manager, suggests a market where advanced AI capabilities are being applied within established operational and governance frameworks.