Key Takeaways
- 1 Visa TAP and Forter’s protocol authenticate AI agents cryptographically, helping merchants distinguish trusted shopping agents from malicious bots.
- 2 OpenAI’s ACP, developed with Stripe, lets customers complete purchases inside ChatGPT, creating a direct path from AI-assisted discovery to checkout.
- 3 Google’s AP2 uses cryptographically signed mandates to record what a user has authorised an agent to buy, creating an audit trail for agent-initiated payments.
Note: This content had corrections applied after publication.
Understanding Agentic Commerce Protocols
Agentic commerce represents a paradigm shift in digital retail, where autonomous AI agents act as intermediaries between consumers, merchants, and payment systems. This model prioritises agent-driven interactions over traditional e-commerce, enabling seamless, real-time decision-making. For global fashion retailers, the adoption of agentic commerce protocols offers transformative potential, while introducing complex operational and technical challenges.
Core Characteristics of Agentic Commerce Protocols
- Autonomy: AI agents operate independently, minimising reliance on human intervention for tasks like product discovery and purchase execution.
- Security: Protocols incorporate cryptographic authentication, fraud detection, and encryption to safeguard transactions.
- Interoperability: Standardised interfaces allow agents to interact with multiple platforms, payment systems, and third-party tools.
- Scalability: Designed for high-volume transactions across global markets, ensuring adaptability to fluctuating demand patterns.
For fashion retailers, these protocols address critical pain points such as customer retention, inventory management, and cross-border logistics. The following sections analyse leading agentic commerce protocols and their strategic relevance.
Overview of Key Agentic Commerce Protocols
1. Visa’s Trusted Agent Protocol (TAP)
Overview: Visa’s Trusted Agent Protocol (TAP) is designed to enhance transparency and security in agentic commerce—where AI agents (e.g., shopping assistants, price comparison tools) interact with merchants. The protocol addresses the critical challenge of distinguishing trusted agents from malicious bots, enabling merchants to engage with these agents securely and effectively. This capability is increasingly vital as AI-driven interactions increasingly shape the customer journey before checkout. Key Features:
- Security: Cryptographic authentication using agent-specific signatures built on the HTTP Message Signature standard.
- Integration: Compatibility with Visa’s existing payment infrastructure.
Relevance to Fashion Retailers: - Reduces fraud risk for high-value fashion items.
- Leverages Visa’s global payment ecosystem to reduce implementation costs.
2. OpenAI’s Agentic Commerce Protocol (ACP)
Overview: ACP, developed with Stripe, enables seamless purchases via AI agents embedded in platforms like ChatGPT.
Key Features:
- Open-Source: Apache 2.0 license for customisation.
- Interoperability: Structured state management for cross-platform interactions.
Relevance to Fashion Retailers: - Supports hyperpersonalised shopping experiences (e.g., AI stylists).
- Provides scalable infrastructure for cross-border transactions, reducing international friction.
3. Google Cloud’s Agent Payments Protocol (AP2)
Overview: Google Cloud’s Agent Payments Protocol (AP2) enables AI agents (e.g., chatbots, virtual assistants) to securely initiate and manage payments within e-commerce workflows. It bridges the gap between conversational AI and payment systems, allowing seamless transactions while maintaining compliance and security.
Key Features:
- Verifiable Mandates: Cryptographically signed mandates record the user’s instructions and approved cart, creating an audit trail for authorisation and accountability.
- Open Standard: An open protocol with a public specification and reference implementations, usable as an extension of the Agent2Agent (A2A) protocol and the Model Context Protocol (MCP).
Relevance to Fashion Retailers: - Enables delegated purchasing, with agents buying autonomously within limits the customer has pre-approved.
- Reduces fraud exposure through authenticated interactions.
4. Forter’s Trusted Agentic Commerce Protocol
Overview: Forter’s Trusted Agentic Commerce Protocol is designed to establish trust and security in AI-driven commerce by verifying the authenticity of AI agents (e.g., chatbots, virtual assistants) interacting with merchants. It aims to prevent fraud, ensure compliance, and enable seamless transactions in an era of increasing agentic interactions.
Key Features:
- Mutual Authentication: Verifies both the agent’s identity and its relationship to the consumer it acts for.
- Encrypted, Bi-directional Data Exchange: Uses JSON Web Encryption to protect shared data, and lets merchants send order, dispute and shipping updates back to the agent.
Relevance to Fashion Retailers: - Essential for high-risk, high-value transactions (e.g., luxury fashion items).
- Ensures regulatory compliance while mitigating revenue loss from fraud.
5. Anthropic’s Model Context Protocol
Overview: While not a commerce protocol, it could be considered a baseline for interacting with commerce agents. Anthropic’s Model Context Protocol (MCP) is an open standard for connecting AI applications such as Claude to external data sources, tools and workflows, such as a retailer’s product catalogue, inventory or order systems. It gives AI assistants one consistent way to retrieve information and take actions, without a bespoke integration for each system.
Key Features:
- Tool and Data Access: A standard interface connecting AI assistants to catalogues, inventory and other business systems.
- Customisation: Tailored for brand-specific interactions.
Relevance to Fashion Retailers: - Enables sophisticated personalised customer service experiences (e.g., AI stylists).
- Improves product recommendation accuracy, boosting conversion rates.
6. Cloudflare’s Web Bot Auth
Overview: Cloudflare’s Web Bot Auth lets bots and AI agents prove their identity by cryptographically signing their web requests, which websites can then verify against the agent’s published keys. In October 2025 Cloudflare extended it to agentic commerce, with Visa’s Trusted Agent Protocol, Mastercard Agent Pay and American Express all building on it, so merchants can tell verified shopping agents apart from malicious bots, a critical capability for protecting e-commerce platforms.
Key Features:
- Agent Authentication: Cryptographically signed requests identify verified bots and AI agents.
- Scalability: Handles high traffic volumes without performance degradation.
Relevance to Fashion Retailers: - Prevents scalping of limited-edition items.
- Protects brand reputation by ensuring fair product access.
7. Perplexity’s Merchant Program
Overview: Perplexity appears to have adopted an approach that is similar to Amazon.com. Through the Merchant Program, retailers can incorporate their products into Perplexity results. Merchants share product data with Perplexity free of charge and can be included in its “Buy with Pro” checkout, but it’s not clear how fraud prevention and returns processing will be managed at scale.
Side-by-Side Analysis of Agentic Commerce Protocols
| FEATURE | Visa TAP | OpenAI ACP | Google AP2 | Forter Protocol | Cloudflare Web Bot Auth |
|---|---|---|---|---|---|
| Security | High (cryptographic auth) | Moderate (open-source) | High (signed mandates) | High (fraud detection) | High (signed requests) |
| Interoperability | Limited (Visa ecosystem) | High (cross-platform) | High (open protocol) | Moderate (customisable) | High (payment system compat) |
| Scalability | High (Visa’s global reach) | High (community support) | High (60+ launch partners) | Moderate (merchant setup) | High (traffic handling) |
| Ease of Integration | Moderate (Visa partnership) | High (API-first) | Moderate (new protocol) | Moderate (customisable) | High (plug-and-play) |
| Cost | Low (open specification) | Low (open-source) | Low (open protocol; integration costs vary) | Variable (custom solutions) | Low (free tier available) |
| Use Case Focus | Fraud prevention | Agentic checkout | Agent-initiated payments | Fraud detection, compliance | Agent authentication |
Strategic Implications for Global Fashion Retailers
1. Enhanced Customer Experience
Agentic commerce protocols enable hyper-personalisation, a critical factor in fashion retail. For example:
- AI Stylists: Agents provide curated outfit recommendations based on user preferences and seasonal trends (OpenAI ACP).
- Real-Time Support: Chatbots connected to retail systems through Anthropic’s protocol deliver instant product information and styling advice, improving engagement.
2. Operational Efficiency
- Delegated Purchasing: Google AP2 lets customers pre-authorise agents to buy autonomously within set constraints, such as a budget, using signed mandates.
- Automated Checkout: OpenAI ACP streamlines the purchase process, letting customers complete checkout without leaving the conversation.
3. Security and Trust
- Fraud Mitigation: Visa TAP and Forter’s protocols are designed to reduce fraud and chargeback risk by authenticating agents before they transact.
- Compliance: Secure protocols ensure adherence to data protection laws like GDPR, critical for global retailers.
4. Global Scalability
- Cross-Border Transactions: OpenAI ACP’s checkout specification carries ISO 4217 currency codes and works with any compatible payment provider, and Google AP2 is designed as a global foundation with further payment methods on its roadmap; both launched with a limited initial scope, so international readiness should be confirmed market by market.
- Local Partnerships: Protocols like Visa TAP leverage existing payment infrastructures, enabling seamless integration in emerging markets.
5. Competitive Advantage
Early adopters of agentic commerce can differentiate themselves by offering innovative shopping experiences. For instance, a fashion retailer using ACP could allow customers to “buy it in ChatGPT,” creating a frictionless path from discovery to purchase.
Challenges and Considerations
1. Security Risks
- Agent Spoofing: Malicious actors could create fake agents to mimic legitimate ones.
- Data Privacy: Handling sensitive customer data requires robust encryption and compliance with regulations like GDPR.
2. Integration Complexity
Adopting agentic commerce protocols often demands technical expertise and infrastructure upgrades. Small to mid-sized retailers may face considerable implementation costs, even where the underlying protocols are open.
3. Regulatory Hurdles
Global fashion retailers must navigate complex, evolving regulatory landscapes:
- EU’s AI Act: Mandates transparency and accountability for AI systems, including agentic commerce platforms.
- Cross-Border Data Transfer: Compliance with data localisation requirements in jurisdictions including China and India.
4. Ethical Concerns
AI agents must be designed to avoid algorithmic bias in product recommendations or pricing strategies. Retailers must ensure fairness, transparency, and inclusivity in their agentic systems to maintain customer trust.
Future Directions
The evolution of agentic commerce protocols will likely focus on:
- Decentralised Identity Systems: Blockchain-based credentials for agent authentication.
- AI-Driven Personalisation: Leveraging large language models (LLMs) to deliver increasingly sophisticated, context-aware shopping experiences.
- Global Standards: Development of unified industry standards to reduce fragmentation and improve interoperability.
For fashion retailers, maintaining competitive advantage will require strategic investments in AI infrastructure and collaborative partnerships with protocol developers and technology providers.
Conclusion
Agentic commerce protocols represent a paradigm shift in how global fashion retailers operate, offering substantial opportunities to enhance customer engagement, streamline operations, and mitigate risks. While challenges such as security and integration persist, the benefits—ranging from hyperpersonalisation to cross-border scalability—present compelling value propositions.
As the industry evolves, fashion retailers must adopt a proactive approach, leveraging agentic commerce to maintain competitiveness in an increasingly AI-driven market. By understanding and strategically implementing appropriate protocols, organisations can achieve new levels of efficiency, innovation, and customer satisfaction.
References:
- Forter. Trusted Agentic Commerce Protocol. https://www.forter.com
- OpenAI. Agentic Commerce Protocol (ACP). https://developers.openai.com
- Google Cloud. Agent Payments Protocol (AP2). https://cloud.google.com
- Anthropic. Model Context Protocol. https://www.anthropic.com
- Cloudflare. Web Bot Auth for Agentic Commerce. https://www.cloudflare.com
- Perplexity. Shop like a Pro. https://perplexity.ai