Personal AI agents, such as Meta's Muse and OpenAI's Dots, are reshaping internet traffic by performing tasks like booking flights or hotel rooms. These agents have recently been deployed at scale, with Muse downloaded 5 million times in its first month. Agentic web traffic tripled from August to September, signaling a shift in internet usage patterns.
Websites are struggling to accommodate these agents, with many blocking them due to security concerns. Platforms like Amazon have blocked Muse's agents and are fighting Perplexity in court, arguing that AI agents threaten their advertising business. TechCrunch reported users having trouble deploying agents to sites like Pizza Hut and Yelp, which restrict AI from booking restaurant reservations.
Ethan Mollick, an associate professor at the Wharton School of the University of Pennsylvania, predicts that as demand for agents grows, platforms will find ways to let them in.
This is especially true for e-commerce sites, where platforms can profit from agents shopping on behalf of real people. Shopify is already welcoming agents by providing them with dedicated checkout lanes that bypass human-designed pages.
"Humans will eventually get very annoyed when their agents aren’t allowed to do this on their behalf," Mollick recently wrote on LinkedIn. He highlights the potential for customer service departments to become overwhelmed as agents request refunds or negotiate prices.
The increasing presence of AI agents could lead to a web optimized for them, making it more difficult for humans to navigate and use the internet as intended.
The future of the internet remains uncertain as rules for agent behavior are still being written.
While some rules will benefit all users, others, like whether businesses can block agents or prevent them from negotiating better deals, will determine whether the web becomes a highway for agents working for us or against us.
The pressing question is what will happen to the internet for the rest of us in the process.
Source: mittr
