World ID wants you to put a cryptographically unique human identity behind your AI agents

In recent months, innovative tools like OpenClaw have demonstrated the capabilities of tech-savvy users leveraging automated agents for various tasks. However, this convenience has raised significant challenges for online service providers, who face overwhelming requests from numerous agents simultaneously, reminiscent of a DDOS attack. To tackle this issue, identity startup World has introduced its 'proof of human' technology called World ID. Today, the company launched a beta version of Agent Kit, a novel approach that allows users to affirm they are controlling their AI agents. This tool also enables websites to restrict access to AI agents by confirming they are operated by actual human beings. World is perhaps best known as the organization behind WorldCoin, a cryptocurrency initiative founded by Sam Altman that began in 2023, offering free coins to users who scanned their irises using a specialized device. While WorldCoin continues to operate, albeit at a value significantly lower than its early highs, World has shifted its focus towards World ID. This technology employs the same iris-scanning method to create a cryptographically secure and unique online identity token, which users can store on their devices. Currently, World claims that almost 18 million individuals have successfully verified their identities through nearly 1,000 physical orbs scattered across the globe. With the introduction of Agent Kit, World aims to empower users to link their verified identities to AI agents, allowing these agents to function on their behalf online in a trustworthy manner. Instead of completely blocking automated traffic as a safety measure, World proposes that websites require AI agents to present a World ID token. This would confirm that a legitimate human is backing each request. This innovative approach could facilitate access for AI agents to limited resources, such as making restaurant reservations, purchasing tickets, accessing free trials, or even managing bandwidth, without the risk of a single user inundating the system with countless anonymous bots. Furthermore, this concept holds promise for protecting sensitive online environments like forums and polls, where preventing automated manipulation is crucial.