Amazon and OpenAI have unveiled a major multi-year partnership aimed at advancing artificial intelligence capabilities for businesses, developers and consumers worldwide, with Amazon committing up to $50bn to support the collaboration.

The funding plan includes an initial $15bn investment, followed by a further $35bn tied to agreed milestones, underscoring a deepening alliance between the companies as global competition in AI infrastructure and deployment intensifies.

At the centre of the agreement is the joint creation of a Stateful Runtime Environment powered by OpenAI models and delivered through Amazon Web Services. The platform is designed to allow AI systems to maintain memory, access computing resources and interact continuously with enterprise software and datasets, enabling developers to run long-term workflows rather than isolated tasks.

The companies described stateful AI environments as the next stage of advanced AI deployment, allowing systems to function persistently within business operations instead of responding only to single prompts.

The platform will integrate with Amazon Bedrock tools and AWS infrastructure services so that AI agents can operate alongside existing enterprise applications. The rollout of the environment is expected in the near term.

As part of the partnership, AWS will also serve as the exclusive third-party cloud provider for OpenAI Frontier, an enterprise platform built to help organisations deploy and manage coordinated teams of AI agents across internal systems while maintaining shared context and strong security controls.

OpenAI Frontier is intended to allow companies to embed AI directly into their workflows without handling complex backend infrastructure, reflecting growing demand for production-level AI deployment rather than experimentation.

The partnership also expands the companies’ broader cloud infrastructure arrangement. OpenAI plans to scale its computing commitments with AWS significantly over the coming years to support rising demand for advanced AI workloads. The agreement includes access to specialised AWS Trainium chips, with newer generations expected to deliver stronger processing power, improved memory performance and greater efficiency for large-scale AI systems.

OpenAI said the expanded capacity will allow it to scale advanced services globally, while AWS customers will gain easier access to sophisticated AI tools delivered on demand through the cloud.

The deal highlights how AI development is increasingly tied to access to massive computing resources and enterprise distribution channels, as companies shift from testing AI applications to embedding them at the core of business operations.