The article examines the impact of AI agents on how we interact with the internet, highlighting a shift from traditional browsing to automated processes. Users might soon rely on AI agents to handle tasks like ordering food or booking tickets, which could significantly increase web traffic from these platforms while decreasing traffic from conventional sources. This change poses challenges for web security, as agents may generate requests at a higher volume and speed, complicating the detection of malicious traffic. Current security measures often block all requests from a source identified as problematic, which could inadvertently affect legitimate users of AI platforms. To address this, the article introduces anonymous credentials (AC) as a potential solution for managing user behavior without compromising privacy. AC allows website operators to enforce security policies and rate-limit requests without tracking individual users. The IETF is developing standards for AC to ensure they work across various platforms. The article also includes a hands-on example of building a simple AI agent that can order pizza, showcasing how the technology can automate browsing tasks. Utilizing tools like Cloudflare’s Browser Rendering service, developers can create agents that navigate websites and perform actions automatically. This opens up possibilities for more sophisticated interactions but also raises concerns about the potential for misuse if agents are given too much control.