AWS makes a series of AI moves including $100 million investment

AWS has announced a further USD $100 million investment in its Generative AI Innovation Centre, expanding its commitment to supporting the development of autonomous and agentic AI systems for customers globally.

This new investment follows two years of the centre working with organisations across the world to improve productivity and transform customer experiences using generative AI solutions. Among the centre's collaborative efforts are projects with Warner Bros. Discovery Sports Europe, which deployed an AI-powered tool for cycling commentators built on Amazon Bedrock and Anthropic's Claude 3.5, and BMW, which uses AWS to enhance diagnostics for over 23 million connected vehicles. Companies such as Syngenta and AstraZeneca have also reported significant benefits from deploying agentic AI technologies in partnership with AWS.

The centre's global team - comprising AI scientists, strategists, and engineers - partners directly with customers and industry peers to address challenges in implementing complex AI systems.

New customisation and developer capabilities

At the AWS Summit in New York, AWS unveiled a set of new features aimed at increasing the usability of its AI models and platforms. The Amazon Nova suite now allows customers to customise models with improved accuracy and flexibility using Amazon Sagemaker AI, as well as SageMaker HyperPod. Security and access control enhancements have been added to the Nova Act SDK, enabling developers to move prototypes to production and create agents capable of interacting reliably within web browsers.

Amazon also launched Amazon S3 Vectors, which introduces native vector support for AI workloads within cloud object storage. This enables customers to lower the costs of storing and querying vector data by up to 90% compared to conventional methods. S3 Vectors is integrated with Amazon Bedrock Knowledge Bases and OpenSearch Service to streamline retrieval-augmented generation (RAG) and vector search processes.

AI infrastructure advancements

AWS introduced Project Rainier, a new AI compute cluster connecting hundreds of thousands of Trainium2 chips across the United States, designed to support large-scale AI workloads and further machine learning research and deployment.

Tools for developers

The release of Kiro, a new integrated development environment, aims to simplify the experience of developing with AI agents. Kiro supports tasks such as running tests and generating documentation, and it also facilitates collaboration between developers and AI agents on project planning and workflow automation.

Model Context Protocol (MCP) resources were also highlighted as new support for developers building in the agentic AI landscape. A local AWS API MCP server includes complete knowledge of AWS's API surface, easing integration, while the AWS Knowledge MCP server provides an always-up-to-date MCP with comprehensive AWS documentation, accessible remotely for greater flexibility.

Enhanced AI model availability

TwelveLabs AI models have been added to Amazon Bedrock, allowing customers to transform video libraries into searchable assets by processing visual, audio, and text components in tandem. AWS is the first cloud provider to offer TwelveLabs models for use with natural language queries, with data protection and security baked into the service.

Collaboration and support for startups

AWS and Meta are collaborating to help startups build AI applications using Meta's Llama small language models. As part of a new programme, up to 30 North American startups can each receive up to USD $200,000 in AWS credits and technical expertise from Meta and AWS, supporting the development of new AI tools and products. Llama 4 models from Meta are now accessible through Amazon Bedrock and SageMaker JumpStart for broader public use.

Simplifying AI agent collaboration

The open source SDK Strands Agents, developed by AWS, has received updates intended to make it easier for businesses to deploy systems where multiple AI agents solve complex problems together. This technology reduces the time required to build co-ordinated AI assistant teams, enabling new applications in customer service, data analysis, and beyond.

Professional development and training

AWS has launched the AWS AI League, a gamified programme where developers compete to solve real-world challenges with generative AI. The initiative makes up to USD $2 million in AWS credits available, along with the chance to win trips to AWS re:Invent and cash prizes. The organisation also offers eight game-based cloud skill training experiences for improved learning outcomes.

To support early-career professionals, AWS is offering students at more than 6,600 AWS Academy institutions free access to advanced training content and certification exam vouchers. Non-Academy students and recent graduates can also benefit from newly published research highlighting promising career paths and Skill Builder learning plans. AWS said it aims to engage with 2.7 million students and early-career professionals globally within the programme's first year.

AWS announced it is making a second $100 million investment in the AWS Generative AI Innovation Centre. The funding builds on two years of the center empowering thousands of customers around the world to boost productivity and transform their customers' experiences.