The Future of AI Commerce: Competing Protocols from Google, OpenAI, and Visa

The landscape of AI-enabled commerce is rapidly evolving, particularly following the announcement of Walmart's integration with OpenAI's ChatGPT. This development raises critical questions about how quickly AI agents can securely complete transactions for consumers.

As chat platforms like ChatGPT increasingly replace traditional browsers, they are becoming adept at providing users with curated product recommendations. Consumers often rely on these platforms for advice on items like the best humidifiers, only to find themselves clicking through to online purchases.

However, a significant barrier remains: AI agents currently lack the trust infrastructure necessary to reassure both consumers and banking institutions about the security of financial transactions. For agents to effectively facilitate purchases, a shared language must be established among model providers, banks, merchants, and consumers.

Emergence of Competing Protocols

Recently, three major players have introduced competing standards aimed at enhancing trust in AI transactions:

• Google’s Agent Pay Protocol (AP2): Developed in collaboration with partners such as PayPal, American Express, and Mastercard, this protocol seeks to ensure secure transactions.
• OpenAI and Stripe’s Agentic Commerce Protocol (ACP): This protocol is designed to minimize infrastructure changes while allowing agents to relay information directly to merchants.
• Visa’s Trusted Agent Protocol (TAP): Launched to provide a secure framework for AI shopping, TAP incorporates a system of approved agents and digital keys for verification.

While these protocols aim to establish a trust layer for AI agents, they also risk creating walled gardens that could hinder interoperability. Enterprises might find themselves locked into a specific platform or payment standard, complicating the shopping experience.

The Challenge of Standardization

Historically, the emergence of multiple standards in any industry can lead to confusion and slow progress. The rapid pace of innovation in AI has made it imperative for companies to adapt quickly, but the presence of three competing protocols may impede this process.

Each protocol has its unique approach to proving authorization. For instance, both AP2 and TAP utilize cryptographic proofs to verify that an agent is acting on behalf of a user. TAP requires agents to be added to an approved list, while AP2 employs a digital contract as a proxy for user approval.

OpenAI’s ACP, on the other hand, operates with less infrastructural demand, allowing for a more streamlined interaction between agents and merchants.

Potential for Walled Gardens

Despite the intention for these protocols to work across various chat platforms, actual interoperability is not guaranteed. Competing standards may create isolated ecosystems, complicating transactions for enterprises that need to trust the agents representing their customers.

Louis Amira, cofounder and CEO of Circuit and Chisel, expressed concerns regarding the future of these protocols, stating, “The better the protocol proposals get, the more likely they are to end up being walled gardens and very hard to interoperate.” This situation could lead to a fragmented marketplace where enterprises struggle to navigate differing standards.

The Path Forward

For now, enterprises are encouraged to experiment with the various protocols and remain hopeful for a consensus that incorporates the best features of each proposal. Wayne Liu, chief growth officer at Perfect Corp, emphasized the importance of open-source solutions in driving this integration.

As the industry continues to explore these new frontiers in AI commerce, the necessity for a universally accepted standard remains paramount. Until trust is established in the ability of AI agents to handle financial transactions, the future of AI-enabled shopping will remain uncertain.

Rocket Commentary

The rapid integration of AI into commerce, exemplified by Walmart's collaboration with OpenAI's ChatGPT, presents both exciting opportunities and significant challenges. While the prospect of AI agents providing tailored product recommendations is promising, the current lack of a robust trust infrastructure poses a serious barrier to consumer adoption. For AI to truly transform commerce, it must not only enhance user experience but also ensure the security and privacy of financial transactions. As we navigate this evolving landscape, industry players must prioritize ethical frameworks and transparent practices to foster consumer confidence and unlock the full potential of AI in commerce. The future hinges on creating a secure, accessible, and trustworthy environment for both consumers and institutions.