AWS

AWS in Singapore:

• Data centres: one of Southeast Asia's largest cloud regions, with multiple Availability Zones
• AI services: Bedrock, SageMaker, Rekognition, Transcribe, and others
• Local partnerships: provides sponsored training compute to the SEA-LION project; cloud-services collaborations with AISG, government agencies, and local banks
• Talent training: local talent programmes such as AWS Academy and re/Start

The USD 9 billion investment announcement (2024) made Singapore the core expansion point for AWS in Southeast Asia.

In Singapore's AI ecosystem, AWS is the "compute and infrastructure layer".

Concrete roles:

• Compute supply: the main cloud provider for local enterprises training and running AI models
• AI services: Bedrock lets enterprises one-click access models from Claude, Llama, Cohere, etc.
• SEA-LION compute: part of the SEA-LION training compute is sponsored by AWS
• Government cloud: parts of Singapore government workloads run on AWS, indirectly shaping AI deployment

AWS does not lead research technically, but the boundaries of its services (which models Bedrock offers, which frameworks SageMaker supports) directly shape local enterprises' AI choices.

The AWS USD 9 billion investment is the single largest funding injection into the "infrastructure lever" of Singapore's AI strategy.

In the "seven transmission levers" framework:

• Lever 1 (infrastructure): the compute base
• Lever 3 (industry adoption): services like Bedrock lower the barrier to enterprise AI deployment

A take: the scale of the AWS investment is a double-edged sword for Singapore — it brings massive capital, jobs, tax revenue, and AI compute, but it also makes Singapore's AI ecosystem heavily dependent on AWS. If Bedrock raises prices or AWS exits an AI service, local enterprises take a serious hit. This is part of the motivation for Singapore pushing SEA-LION (a localised model that does not depend on AWS APIs).