Akamai & Visa join forces to secure AI shopping

Akamai has formed a strategic collaboration with Visa that links Visa's Trusted Agent Protocol with Akamai's edge-based security services for AI-driven online shopping.

The partnership focuses on identity, user recognition, and fraud controls for so-called agentic commerce, where autonomous AI systems browse and buy on behalf of consumers. The companies plan to support merchants that expect a rapid increase in AI agents interacting with their websites.

Under the arrangement, Visa's Trusted Agent Protocol will run across Akamai Cloud. Akamai describes its platform as the world's most distributed cloud infrastructure for security and content delivery.

AI agents as shoppers

Agentic commerce refers to digital shopping activity carried out by AI agents operating on behalf of human users. These agents can search for products, compare prices, and initiate purchases without direct real-time input from the consumer.

The rise of such agents introduces a new category of automated traffic that differs from traditional web crawlers and bots. Merchants need to distinguish between legitimate AI agents with clear authority from a consumer and malicious automated traffic that aims to abuse systems.

Visa's Trusted Agent Protocol offers an authentication framework for AI agents that use Visa credentials for payments. Akamai will add behavioural and network intelligence, user recognition, and bot and abuse detection at the network edge.

Through this approach, the firms intend to give merchants near real-time visibility into AI agent activity before it reaches payment systems or sensitive applications. This is expected to support risk-based decisions on whether to allow, challenge or block a given interaction.

Dual identity challenge

The collaboration seeks to address what both companies describe as a "dual-identity" problem in AI commerce. Merchants must be able to verify both the agent that is performing the transaction and the human user that the agent represents.

"The promise of agentic commerce hinges on recognition: the fundamental ability to trust an agent acting on someone's behalf," said Patrick Sullivan, Chief Technology Officer, Security Strategy, Akamai Technologies. "By combining Visa Trusted Agent Protocol with Akamai's deep user recognition and threat intelligence, we're working to solve the dual-identity challenge that's crucial to AI commerce. We prove both who the agent is and, critically, who it represents. This is what transforms AI agents from novelties into trusted economic actors."

Akamai plans to use its existing edge-based user recognition tools to preserve the identity of an AI agent once it first appears. The company aims to maintain risk posture, trust signals, and account context as the agent moves across different pages, sessions, or devices.

Visa's Trusted Agent Protocol will allow an AI agent to transmit structured information to a merchant. This information can confirm that the agent is approved for a specific shopping task, identify the consumer on whose behalf it is operating, and present payment credentials in the format the merchant expects.

Fraud risk and bot traffic

The launch comes against a backdrop of rapid growth in AI-driven automated traffic. Akamai's 2025 Digital Fraud and Abuse Report found that AI-powered bot traffic rose 300% over the past year. The report said merchants in the commerce sector recorded more than 25 billion AI bot requests in a two-month period.

In this environment, merchants face a widening attack surface. Attackers can use AI to simulate legitimate shopping behaviour at scale. That increases the difficulty of telling apart acceptable automation from fraud attempts such as credential stuffing, account takeover, or card testing.

The companies say that Trusted Agent Protocol will only mark an agent as trusted if it meets authentication and authorisation checks and if it uses a Visa credential within defined parameters. Akamai will then apply bot detection and behavioural analysis policies around that agent's traffic.

Merchant use cases

Akamai and Visa have outlined three main merchant use cases. The first involves identifying a legitimate AI agent and its intent. Trusted Agent Protocol can signal whether an agent is browsing or attempting to pay. Akamai layers real-time behavioural and network intelligence on top of this to spot unusual patterns.

The second use case links the agent to the underlying user. Trusted Agent Protocol enables agents to pass information that connects each verified agent to the consumer. Akamai's user recognition tools then maintain this linkage at the edge of the network.

The third use case covers payment interactions. Trusted Agent Protocol supports the way agents pass payment details, including network tokens or micropayment flows. Akamai applies what it describes as end-to-end protection on these sessions, verifying that the agent remains consistent with its earlier authentication and that the transaction does not show signs of abuse.

Visa says Trusted Agent Protocol relies on standard web infrastructure. The payments company positions this as a way for merchants to adopt the system without extensive reconfiguration of checkout flows or user interfaces.

Global reach

Visa works with around 175 million merchant locations globally. Adoption of Trusted Agent Protocol within this network could give the standard a wide footprint across card-not-present commerce if large retailers implement it.

Nine of the world's top ten retailers already use Akamai for digital commerce services, according to the company. That reach spans peak traffic management, security controls and performance optimisation for eCommerce sites.

Visa describes itself as a global digital payments leader. The company operates across more than 200 countries and territories and handles transactions for consumers, merchants, financial institutions and public sector entities.

The two firms intend to align their respective infrastructures so that merchants can engage with authorised AI agents at scale while maintaining a consistent set of trust signals.

In a statement on the broader industry context, Visa said merchants and other ecosystem participants need a common way to trust AI agents.

"Agentic commerce is unlocking an entirely new wave of digital interactions, but it can only scale if every player in the ecosystem can trust the agents participating in it," said Jack Forestell, Visa Chief Product & Strategy Officer. "By collaborating with Akamai to deploy Trusted Agent Protocol, we're delivering the real-time intelligence merchants need to support AI-driven experiences without introducing new risk. This is how we help the industry move confidently into the next era of commerce."