
Meta, Walmart, Stripe and other companies are launching an open standard called the 'personal agent protocol' to regulate how AI agents interact with businesses, aiming to improve authentication, security and transparency as Meta's Muse agent gains millions of users.
AI-generated summary
Meta launched Muse, a personal AI agent, a month prior to this announcement. Muse became a viral sensation, reaching the top of Apple's App Store. Amazon has blocked Meta's agents and Perplexity's AI agent over scraping concerns, even suing Perplexity in November.
A month after Meta's launch of Muse, the personal agent that quickly turned into a viral sensation, a group of companies is coming together to try to simplify the process of bringing the technology into the business world.
Meta, Walmart , Stripe and a handful of other companies are publishing what they're calling a "personal agent protocol." It's an open standard meant to dictate how artificial intelligence agents can interact with businesses.
Enterprise AI startup Sierra, co-founded by former Salesforce co-CEO Bret Taylor, is part of the group. Taylor, who's also chairman of OpenAI and is leading the initiative, told CNBC in an interview 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. 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," Taylor said. "It is kind of chaos until such a standard exists."
Meta's Muse has become the most popular and buzzy versions of personal AI agents since hitting the market in early September. It soared to the top of Apple's App store and remains there, ahead of ChatGPT. However, Amazon is among the companies that have blocked Meta's agents, citing worries of website scraping.
Amazon also blocked Perplexity's AI agent and sued the startup in November, alleging the company 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."
For the new standard, Taylor compared it to the experience of logging into other websites with Google or Facebook credentials, technology he worked on when he was tech chief at Facebook. He said the practice will help with authentication so that businesses know if they're dealing with a bot, and will provide visibility into what personal agents are actually doing through these websites.
David Singleton, VP of engineering and consumer products at Meta Superintelligence Labs and former technology chief at Stripe, highlighted security. In order for agents to work well, he said, customers need to share credit card and personal information, and the new 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."
Singleton said Muse already has millions of users in the U..S and it's "growing rapidly."
"What we're finding is that people love using Muse for all kinds of things, things like signing their kids up for classes, finding and buying the best gifts, scheduling that dentist appointment that you have been putting off for six months," he said. "We need a mechanism to have it go faster and smoother."
While OpenAI and Anthropic aren't on board now, Taylor said he expects the AI giants to participate, and indicated he will be "really disappointed" if it's not being used by competitors.
"The whole point of this is to be an open standard," Taylor said.
AI outlook — possibilities, not facts
OpenAI and Anthropic will eventually join the personal agent protocol initiative
Likely · Within months
The personal agent protocol will see initial adoption by businesses within the next 3-6 months
Possible · Within months

Samsung's 2025 65-inch The Frame TV is discounted to $898, a 40% reduction from its original $1,498 price, making it the lowest price ever for the popular art TV model. The deal highlights the TV's matte screen for art display, brightness for well-lit rooms, and limitations in contrast and free art access, while noting the newer Frame Pro (2026) is also on sale but less competitively priced.

Hark, a startup founded by Brett Adcock, launches Hark Pro, an AI personal assistant designed to perform computer tasks like managing emails and booking flights. The product emphasizes transparency by showing the agent's web navigation and aims to build an AI-native operating system and device by 2027, distinguishing itself from competitors through task-specific training and user trust.

Dozens of ASOS app users in the UK received pop-up messages claiming hackers had compromised the company's Snowflake instance and demanding engagement or threatening data leaks. The messages, sent via the app's notification system, included a link to a newly created Telegram channel called Xuanye Group. Cybersecurity experts noted the attack's unusual public nature and suggested access may extend beyond Snowflake to ASOS's notification systems.

Mistral AI launched Mistral Large 4, a 1-trillion-parameter AI model using a mixture-of-experts design with 49 billion active parameters per query. The model is positioned as a competitive open-weight alternative to Claude Opus 5.5 and GPT-6 Astra, with pricing at $1.36 per million input tokens and $4.18 per million output tokens. Benchmarks show strong performance in coding and automation tasks, trailing only Kimi K3 and Gemini 4 Argon in some tests. Mistral plans to release the model weights by end of October.

Benchmark is leading a $25 million seed round into Furientis, a one-year-old defense startup building low-cost interceptor missiles to address U.S. military stockpile shortages. The company has already secured a Pentagon contract and is valued at $125 million, reflecting renewed VC interest in defense tech following conflicts in Ukraine and Iran.

In his statement to Politico, OpenAI CEO Sam Altman stated his differences of opinion with Anthropic CEO Dario Amodei on artificial intelligence regulations. Altman emphasized that the benefits of artificial intelligence require accepting some risks and that he does not adopt an approach that aims to completely eliminate the risks of misuse. Anthropic announced that its regulatory recommendations only cover leading artificial intelligence models.