Meta launched its personal agent Muse a month later, and this product quickly became popular. Now, a group of companies are joining forces to try to simplify the process of introducing this technology into the business world.
Meta, Walmart, Stripe, and several other companies are releasing a standard they call the "Personal Agent Protocol." This is an open standard designed to regulate how artificial intelligence agents interact with businesses.
Enterprise AI and startup Sierra are also members of this organization. Sierra was co-founded by Brett Taylor, former co-CEO of Salesforce. Taylor also serves as the chairman of OpenAI and is leading this initiative. In an interview with CNBC, he stated that there is still a lot of work to be done to determine when and how individual agents can access information.
Taylor said, "Companies will be able to know whether they are dealing with a personal agent or a real person. For many companies, there is a risk involved: you don't want a random robot that is not acting on behalf of anyone to gain access to this service." He also stated, "Until such standards are established, the situation is somewhat chaotic."
Since its launch at the beginning of September, the Muse of Meta has become the most popular and highly discussed version among personal AI proxies. It has topped Apple's App Store rankings and continues to hold that position, leading ahead of ChatGPT. However, Amazon is one of the companies that block Meta proxies, due to concerns about website content being crawled.
Amazon also blocked the AI proxies of Perplexity, and in November sued this startup, accusing it of taking measures to “hide” its AI proxies in order to continue crawling the Amazon website without approval. Perplexity called this lawsuit a “bullying tactic.”
When talking about this new standard, Taylor compared it to the experience of logging into other websites using an account with Google or Facebook. This technology is also something he was involved in when he served as the technical lead at Facebook. He stated that this approach will help with authentication, allowing companies to determine whether they are dealing with robots, and it will also enable them to see what individual proxies are actually doing on these websites.
David Singleton, the Vice President in charge of engineering and consumer products at the Super Intelligent Laboratory, who is also the former technical director of Stripe, emphasized security. He stated that for agents to be truly useful, customers need to share their credit card and personal information, and this new standard will bring a certain degree of visibility and control.
Singleton said, "We are defining a set of standards that we hope both individual agents and corporate agents will be able to operate on in the future." He compared it to email, saying, "If you think about email, it works well because it is a standard that everyone can use to communicate with each other. We believe that this will be very similar as well."
Singleton indicates that Muse already has millions of users in the United States, and is "growing rapidly."
He said, "We've noticed that people like to use Muse for various tasks, such as registering their children for courses, finding and purchasing the most suitable gifts, or scheduling dental appointments that they've been putting off for six months. We need a mechanism to make all of this happen faster and more smoothly."
Although OpenAI and Anthropic have not yet joined, Taylor stated that he expects these AI giants to eventually get involved, and said that he would be "very disappointed" if competitors do not adopt this standard.
Taylor said, "The whole point of this is that it must be an open standard."
