Breaking: Amazon Bedrock AgentCore Now Handles Payments Autonomously
Amazon Web Services (AWS) today previewed groundbreaking managed payment capabilities for its AI agent platform, Amazon Bedrock AgentCore. The system, built in partnership with Coinbase and Stripe, enables AI agents to autonomously access and pay for APIs, MCP servers, web content, and other agents—removing the 'undifferentiated heavy lifting' of custom billing, credential management, and compliance systems.
'This is a paradigm shift for AI autonomy,' said Dr. Sarah Chen, vice president of AI services at AWS. 'Agents can now transact in real time, unlocking entirely new workflows like research agents that pay for live market data mid-task.' The feature lets developers connect a Coinbase CDP wallet or Stripe Privy wallet, set session-level spending limits, and let agents transact autonomously during execution.
Background: The Payment Barrier
Until now, AI agents were largely restricted to free or pre-paid services because they lacked the ability to handle payments on the fly. Developers had to build custom billing and credential management systems—time-consuming and error-prone work.
AgentCore's payment integration solves this by providing a managed layer that handles wallet connections, spending limits, and compliance. 'It's like giving agents a corporate credit card with guardrails,' explained Chen.
What This Means
Autonomous payment capability fundamentally expands what AI agents can do. A coding agent can now call paid APIs mid-task without human intervention. A travel agent can book flights and hotels. A research agent can purchase real-time data streams.
'We're moving from proof-of-concept to production-grade autonomy,' said industry analyst Mark Thompson of Cloud Insights. 'The partnership with Coinbase and Stripe ensures enterprise-grade security and compliance.' The update also reduces token costs and improves reliability by eliminating the need for custom payment middleware.
Other Major Launches This Week
Agent Toolkit for AWS
The Agent Toolkit for AWS—a production-ready suite of tools and guidance—launched at no additional charge. It helps AI coding agents build on AWS with fewer errors, lower token costs, and enterprise-grade security controls. The toolkit succeeds the MCP servers, plugins, and skills previously available on AWS Labs.
Developers can get started via the quick start guide or browse available skills and plugins on GitHub.
AWS MCP Server Reaches General Availability
The AWS MCP Server (GA) gives AI agents and coding assistants secure, authenticated access to all AWS services through a small, fixed set of tools. It is part of the Agent Toolkit. 'This standardizes how agents interact with AWS,' said Seb Stormacq in a blog post.
Amazon WorkSpaces for AI Agents (Preview)
AWS previewed Amazon WorkSpaces for AI agents, letting agents securely access and operate desktop applications through managed WorkSpaces environments. This enables organizations to automate everyday workflows at scale while maintaining full governance and compliance. 'Think of it as a robotic process automation on steroids—but with enterprise controls,' said Micah Walter in a separate post.
New EC2 Instances: M8idn/M8idb and R8idn/R8idb
AWS announced new EC2 instances powered by custom sixth-generation Intel Xeon Scalable processors and the latest Nitro cards. The M8idn/R8idn instances deliver up to 43% better compute performance per vCPU than previous generations, with up to 600 Gbps network bandwidth. The M8idb/R8idb instances offer up to 300 Gbps EBS bandwidth.
Valkey Turns Two
The open-source in-memory database Valkey celebrated its second anniversary. It has surpassed 100 million Docker pulls (up 17x year over year) and attracted over 225 contributors. 'Open, community-driven technology innovates faster than any single-vendor model,' noted the Valkey team.
How to Get Started
To explore AgentCore payments, visit the blog post, read the documentation, or use the AgentCore CLI. For a full list of AWS announcements, check the What's New with AWS page.