Meta, Walmart, Stripe Push Open Standard for AI Agents as Websites Block Bots

Bret Taylor, co-founder of Sierra, chairman of OpenAI and former Salesforce co-CEO, is leading the initiative. He told CNBC that companies have a lot of work to do to figure out how and when personal agents are able to access information. “Companies will know when it’s a personal agent versus an actual person,” Taylor said. “For a lot of companies there’s a risk: you don’t want just a random bot that isn’t acting on behalf of a person to have access to this service. It is kind of chaos until such a standard exists.”

Meta’s Muse launched in early September and became one of the most popular and buzzy personal AI agents, soaring to the top of Apple’s App Store, where CNBC reported it remains ahead of ChatGPT. David Singleton, head of Meta Superintelligence Labs and former technology chief at Stripe, said Muse already has millions of users in the United States and is growing rapidly. He said people use Muse for tasks such as signing their children up for classes, finding and buying gifts, and scheduling dentist appointments. “We need a mechanism to have it go faster and smoother,” Singleton said.

Taylor compared the new standard to logging into websites with Google or Facebook credentials, technology he worked on when he was technology chief at Facebook. He said it will help with authentication so businesses know whether they are dealing with a bot, and will provide visibility into what personal agents are doing through websites. Singleton highlighted security, saying customers need to share credit card and personal information for agents to work well, and the standard creates some level of visibility and control. “We’re defining rails that we hope personal agents and business agents can run over for the future,” Singleton said, comparing it to email. “If you think about email, it works great because it is a standard that everyone can use to talk to each other, and we think this is going to be quite similar.”

The push follows growing friction between AI agents and websites. Amazon has blocked Meta’s agents, citing worries of website scraping, and CNBC reported that Amazon also blocked Perplexity’s AI agent and sued the startup in November, alleging it took steps to “conceal” its AI agents so they could continue to scrape Amazon’s website without approval. Perplexity called the lawsuit a “bully tactic.”

TechCrunch reported that consumers adopting agents like Muse, Instinct, ChatGPT’s Dots and others often run into issues when a website on the other end blocks the AI from completing a job. Amazon recently began blocking Muse from its retail site, meaning the agent could no longer browse or make purchases from its product catalog. The problem is not limited to Amazon or Muse, according to TechCrunch, and user complaints across social media indicate agents often encounter similar blocks. Some blocks are intentional, while others result from traditional anti-bot measures designed to protect against spam and malicious activity. End users often cannot tell whether a block is accidental, leaving them frustrated with both the service provider or retailer and the agent.

Walmart illustrates the confusion. TechCrunch saw complaints on social media claiming that Muse could not complete a purchase on Walmart’s retail website; NBC also reported that this was a problem Walmart had faced in its tests of the AI agent. But a Walmart spokesperson told TechCrunch the blocks were not intentional. Walmart is partnered with Muse, as announced at Meta’s Connect developer conference in September, and the company said it wants to be where its customers are, including through AI agent experiences. The issue often appears to involve a button that must be clicked to verify the visitor is human. If that experience is interrupted, the check can fail and the agent gets booted out. TechCrunch said this indicates current technology used to verify humans and keep websites safe from spam and bots may not fit the agent-first era of the web.

To address the problem, industry partners including Meta, Walmart, Stripe, Sierra, Genesys, Rocket, NiCE and Decagon have begun working on an open standard that would dictate how AI agents can communicate with businesses, helping separate good bots acting on behalf of users from bad ones. Meta said in a blog post that the protocol will focus specifically on agent-to-agent communication for online commerce.

Other companies have not fully embraced the technology. Flight booking is one of the main use cases positioned for agents, but many people have reported issues with airlines rejecting their agents’ requests. Delta told TechCrunch it is taking steps to protect customers from unauthorized automated activity and that if it were to open up to AI agents, it would have to be done with “security and the customer experience in mind.” Heena Chavda, Delta’s general manager of global communications, said Delta does not currently have a partnership or integration enabling a third-party AI agent to shop for or book Delta flights on a customer’s behalf on its digital platforms, and that it continues to evaluate how new technology, including external AI agents, could enhance customer engagement. United pointed TechCrunch to its terms of use, which say a user agrees “not to use any robot, spider, other automatic device, or manual process to monitor or copy this website or the content contained therein or for any other unauthorized purpose without United’s prior written permission.” United did not respond to a follow-up question about whether it was blocking specific personal AI agents such as Muse, according to TechCrunch.

Early adopters’ complaints on social media have referenced a number of brands reportedly blocking AI agents, including Yelp, eBay, Zillow, Pizza Hut, Adidas and various airlines. Some users said they believed blocking activity had recently increased. A couple of people said eBay suspended their accounts because of their use of agentic AI, and another complained that Google kicked out Instinct while it was cleaning up their Gmail inbox. Yelp told TechCrunch it does not permit non-human traffic on its platform unless the agent has paid to access Yelp’s content and data.

OpenAI and Anthropic are not on board now, Taylor said, but he expects the AI giants to participate and indicated he would be “really disappointed” if the standard is not used by competitors. “The whole point of this is to be an open standard,” Taylor said.