Competing Agent Protocols Create Standards Battle in AI Commerce
The race to dominate AI-powered commerce intensified recently as three major players unveiled competing protocols for enabling AI agents to conduct secure transactions, according to industry reports. Google launched its Agent Pay Protocol (AP2), while OpenAI partnered with Stripe on the Agentic Commerce Protocol (ACP), and Visa introduced the Trusted Agent Protocol (TAP), sources indicate.
These developments come as chat platforms increasingly replace traditional browsers for product discovery and shopping. When users ask ChatGPT for product recommendations, they currently must click through to complete purchases manually because AI agents lack the trust infrastructure to handle payments directly, analysts suggest.
Three Approaches to Solving the Trust Problem
The competing protocols all aim to provide the necessary trust layer to convince financial institutions and consumers that AI agents can safely handle transactions, according to reports. However, each takes a distinct technical approach to authorization and security.
Sources indicate that both Visa’s TAP and Google’s AP2 rely on cryptographic proofs to demonstrate an agent is acting on a user’s behalf. TAP reportedly adds agents to an approved list with digital keys, while AP2 uses digital contracts as proxies for human approval. OpenAI’s ACP, meanwhile, functions more as an information courier to merchants without requiring significant infrastructure changes, the report states.
Interoperability Concerns and Walled Gardens
Industry experts express concern that competing standards could create walled gardens where protocols only work on specific platforms, potentially locking enterprises into particular ecosystems. Louis Amira, CEO of agent commerce startup Circuit and Chisel, told VentureBeat that “the better the protocol proposals get, the more likely they are to end up being walled gardens and very hard to interoperate.”
This fragmentation contrasts with the open nature of the traditional web, where Google Chrome, Safari, and other browsers can access the same websites thanks to universal standards like TCP/IP. With AI chat platforms, users typically remain within single ecosystems like OpenAI‘s ChatGPT or Google’s Gemini for their shopping queries.
Enterprise Adoption Challenges
Enterprises now face the complex decision of which protocol to support, with the risk of choosing a standard that might not achieve broad industry adoption. Companies serving as merchants of record must trust that agents contacting them genuinely represent customers, according to industry analysis.
The situation mirrors previous technology standards battles, where multiple proposals typically emerge before the industry coalesces around a winner. However, sources suggest the rapid pace of artificial intelligence innovation has accelerated this process compared to historical standards development.
Path Forward: Experimentation and Potential Convergence
For now, industry experts recommend that enterprises experiment with all available protocols while hoping for eventual convergence. Wayne Liu of Perfect Corp. suggested that “the importance of open source exists because it will be the driving force to put everything together,” according to reports.
The emergence of these protocols follows significant industry moves, including Walmart’s integration with ChatGPT and Salesforce‘s substantial investments in automation. As Visa Inc. and other payment giants enter the arena, the battle for AI commerce dominance appears poised to intensify throughout 2024, with potential for additional players to introduce competing standards.
This article aggregates information from publicly available sources. All trademarks and copyrights belong to their respective owners.
