Google has announced its Universal Commerce Protocol (UCP), a new specification designed to support agentic commerce across the entire shopping journey from product discovery and purchasing to post-purchase support.

UCP is initially being rolled out across Google-owned platforms including Search AI Mode and the Gemini App, with availability beginning through select US retailers.

This announcement reflects a broader shift in digital commerce. AI systems are no longer limited to assisting with product discovery. Instead, they are evolving into end-to-end shopping companions, enabling users to move from intent to purchase without breaking context.

As agentic commerce gains momentum, a natural comparison emerges between Google’s Universal Commerce Protocol and OpenAI’s Agentic Commerce Protocol (ACP). While both aim to reduce friction and deliver seamless AI-driven shopping experiences, they differ in execution, environment, and the role merchants play.

Understanding these differences is critical for businesses preparing for the next phase of commerce.

A Common Objective, Multiple Approaches

Agentic commerce is not converging into a single universal standard.

Instead, multiple protocols are emerging in parallel, each addressing different stages of the buyer journey including discovery, consent, checkout, payment and post-purchase support.

No single protocol currently covers every scenario end to end. However, the shared goal remains consistent:
 to minimize friction between discovery and conversion while enabling AI systems to act on user intent in a way that feels seamless and trustworthy.

Two of the most prominent approaches today are OpenAI’s ACP and Google’s UCP.

OpenAI’s Agentic Commerce Protocol (ACP)

OpenAI’s Agentic Commerce Protocol (ACP) is an open, cross-platform protocol that enables shopping and payments directly within AI assistants.

ACP is designed for broad adoption, independent of any specific interface, platform or distribution surface. This allows AI-led shopping experiences to function consistently across different environments.

Key Capabilities of ACP

ACP defines how AI agents can:

  • Discover products using merchant-provided product feeds
  • Present accurate pricing, inventory, and availability
  • Initiate checkout on behalf of the user without redirecting them away from the experience

Checkout within ACP is powered by delegated payment tokens that are single-use, time-bound, and amount-restricted. This structure balances user control with merchant confidence.

Merchant Ownership and Control

A defining feature of ACP is that it preserves the merchant’s traditional role. Settlement, refunds, chargebacks and compliance remain with the merchant, mirroring established ecommerce models.

ACP supports:

  • Checkout journeys inside multiple AI chat interfaces
  • Payment flows aligned with existing merchant risk and operational frameworks

While ACP offers long-term flexibility and optionality, it requires deeper upfront integration. Merchants must support product feeds and agent-initiated checkout to fully participate.

OpenAI’s Instant Checkout within ChatGPT represents the first live implementation of ACP, distinct from the protocol itself.

Google’s Universal Commerce Protocol (UCP)

Google presents Universal Commerce Protocol (UCP) as a new open standard for agentic commerce, created to establish a shared language that allows agents and systems to work together across consumer platforms, businesses and payment providers.

UCP is described as compatible with existing standards, including:

  • Agent Payments Protocol (AP2)
  • Agent-to-Agent (A2A)
  • Model Context Protocol (MCP)

How UCP Is Being Deployed

At launch, UCP is implemented within Google-owned environments, including:

  • Search AI Mode
  • Gemini App
  • Google Shopping

UCP enables a new checkout experience that allows users to complete purchases from eligible US retailers during product discovery, without leaving Google. Payments are completed using Google Pay, with support for PayPal planned.

Retailers remain the seller of record and can customize their integrations, benefiting from reduced checkout friction and improved conversion potential.

Business Agent: Google’s Conversational Commerce Layer

Alongside UCP, Google introduced Business Agent, a branded conversational experience that allows retailers to interact directly with shoppers within Search.

Business Agent enables retailers to:

  • Respond to customer questions in their own voice
  • Engage users during high-intent shopping moments
  • Support direct purchases, including agentic checkout

This experience appears to be a Google-surface implementation aligned with the same agentic commerce stack, supporting conversational discovery and transactional capabilities.

UCP Today and What’s Next

At launch, UCP-powered experiences focus primarily on:

  • Checkout and payments
  • Order-related workflows
  • Conversational discovery within Google surfaces

Google has indicated that future enhancements will include capabilities such as discovering related products, applying loyalty rewards and enabling custom shopping experiences.

For now, UCP’s first execution remains centered within Google’s ecosystem even as it positions itself as an open standard.

How ACP and UCP Coexist

Google has stated that UCP is compatible with existing agentic commerce protocols and agent-to-agent standards. While full interoperability details have not been publicly defined, the positioning suggests that UCP is intended to coexist with open protocols like ACP, rather than replace them.

In practical terms:

  • ACP enables agent-led commerce across AI assistant ecosystems
  • UCP focuses on reducing friction and increasing conversions within Google’s platforms

Each protocol addresses different moments of intent and different engagement environments.

Do Businesses Need to Choose a Single Model?

No.

This evolution is not about selecting a single winning protocol. Instead, it’s about understanding which model serves which type of customer interaction.

Google’s approach supports high-intent shoppers already engaging through Search and Shopping. Open, agent-led protocols like ACP enable new demand, where AI assistants become commerce destinations in their own right.

Each model delivers value at different points in the customer lifecycle.

For businesses, the priority is not choosing one over the other but building readiness for both as agentic commerce continues to mature.

Preparing for the Next Phase of Commerce

Rather than attempting to optimize for every emerging protocol simultaneously, businesses should focus on:

  • Strengthening performance across existing digital channels
  • Building flexible commerce and technology stacks
  • Staying informed about evolving agentic commerce standards

This approach ensures adaptability as AI-driven commerce models expand across platforms and user experiences.

Turning AI-Driven Discovery Into Business Growth

Agentic commerce opens new pathways to engage high-intent users, but it also introduces complexity across discovery, consent and checkout flows.

For organizations navigating this shift, understanding how protocols like ACP and UCP shape AI-led shopping experiences is essential to staying competitive in a rapidly evolving digital landscape.