AWS Machine Learning Blog→ original

AWS представила платежи для ИИ-агентов — интеграция Bedrock AgentCore с Coinbase и Stripe

AWS представила новую функцию Amazon Bedrock AgentCore Payments, позволяющую ИИ-агентам самостоятельно совершать платежи в интернете. Партнёрство с Coinbase и S

AWS представила платежи для ИИ-агентов — интеграция Bedrock AgentCore с Coinbase и Stripe
Source: AWS Machine Learning Blog. Collage: Hamidun News.
◐ Listen to article

Amazon Web Services has introduced Amazon Bedrock AgentCore Payments — a new set of features that enables AI agents to autonomously make payments for the services they use. The feature is in preview mode and was developed in collaboration with cryptocurrency exchange Coinbase and payment system Stripe — two of the most influential financial technologies today.

What is AgentCore Payments

Amazon Bedrock AgentCore is an AWS cloud service that helps developers create, manage, and scale AI agents. Previously, agents running on Bedrock operated as semi-autonomous systems: they could execute complex tasks, call external APIs, analyze large volumes of data, and make decisions. But one thing was missing — financial independence. All payments were controlled and approved by humans.

AgentCore Payments changes this fundamentally. The new feature adds a critically important capability to agents: they can now not only execute tasks but also pay for them independently. If an agent requests a paid service — whether it's a commercial API, cloud computing resources, data storage, or an analytics tool — it can initiate a payment and complete the operation without any human intervention. Payment happens instantly, with financial flows conducted through secure channels using modern encryption.

Partnership with Financial Services

The integration with Coinbase and Stripe is not just a business deal, but a strategic choice by AWS. The company spans two fundamentally different, yet equally important segments of the financial system. Stripe handles traditional payments through credit and debit cards, bank accounts, and digital wallets. Coinbase adds support for blockchain, crypto assets, and decentralized financial protocols.

This partnership demonstrates a fundamentally new approach by AWS to cryptocurrencies. Previously, many cloud platforms treated crypto as experimental or marginal technology. AWS does differently: Coinbase receives a place in the core production infrastructure for AI agents — on equal footing with Stripe. This signals that in the AWS ecosystem, cryptocurrencies are becoming a full-fledged option for funding agent operations.

  • Funding operations through traditional payment systems (Stripe, cards, banks)
  • Support for crypto payments and decentralized assets (Coinbase, blockchain)
  • Instant operations available 24/7 without bank holidays
  • Built-in security, encryption, and regulatory compliance

Why It Matters and Where It's Applied

Previously, AI agents operated under a serious limitation: they could execute tasks but couldn't independently pay for the resources they needed. Developers had to manage payments manually, which slowed down workflows, increased development costs, and required constant human oversight of financial operations.

With AgentCore Payments, the cycle is complete. The agent becomes a fully autonomous entity in economic terms. Now it can assess the cost of the service it needs, initiate a payment from the user-allocated budget, and complete the operation — all without intermediate approvals. This not only makes AI agents more independent but enables the automation of complex multi-step business processes in their entirety.

What It Means

Payments for AI agents represent a turning point in their evolution. As long as agents couldn't make payments, they were limited to only free services and pre-paid resources. Now payment capabilities open up access to the entire ecosystem of commercial APIs and cloud services. This doesn't just make life easier for developers. It creates an entirely new class of business models where AI agents become independent economic actors. They can participate in value chains that previously required human approval at every step. In the long term, this means the boundaries between "agent" and "full economic actor" will begin to blur.

ZK
Hamidun News
AI news without noise. Daily editorial selection from 400+ sources. A product by Zhemal Khamidun, Head of AI at Alpina Digital.
What do you think?
Loading comments…