The rise of AI bots, deepfakes and autonomous agents is creating a new challenge for online platforms, according to Tiago Sada, chief product officer at Tools for Humanity. In a recent conversation on The Neuron Podcast, Sada argued that the web is entering a human verification crisis, where it is becoming harder to tell whether a user is a real person or an automated system.
Sada said that issue is not limited to spam or obvious bot traffic. As AI tools become more capable, he suggested that digital systems need a stronger way to confirm that a person is both human and allowed to act on a platform. That is where World ID, the identity product associated with Tools for Humanity, comes in.
The discussion centered on why current methods such as CAPTCHAs and know your customer, or KYC, processes are no longer enough on their own. Sada described these approaches as increasingly fragile in the face of advancing automation. CAPTCHAs are designed to distinguish humans from bots, but they can be bypassed or become an inconvenience for users. KYC systems can confirm identity, but they are not always suited to everyday online interactions and can create barriers to access.
According to the conversation, the problem is not only technical. It is also about trust. Platforms need to know when a user is a person, when that person is unique, and when they are authorized to take an action. Sada framed those as separate questions that the internet currently struggles to answer reliably.
Sada described World ID as a possible trust layer for the internet, one that could help websites and services verify human users without relying only on methods that are either easy to game or too cumbersome for regular use. The goal, as presented in the interview, is to create a system that can identify a user as human while preserving the ability to interact online at scale.
The conversation also touched on how this approach might affect both users and developers. For users, a more reliable human verification system could reduce friction in everyday online activities. For builders and platform operators, it could provide a way to limit abusive automated activity while still enabling legitimate use cases.
Sada also addressed the difficulty of detecting AI-generated activity, saying it can quickly turn into an arms race. As detection tools improve, so do the methods used to evade them. That dynamic, he suggested, makes purely detection-based approaches less durable over time.
The podcast framed the issue as part of a wider shift in how the internet works. If AI agents become more common, the online world may need systems that distinguish not just between humans and machines, but between different kinds of machine behavior. Some of that activity may be useful, while other forms may be abusive.
That distinction was reflected in the episode’s structure, which included discussion of good agents versus abusive bots and what World ID actually is. The broader message was that identity and verification are becoming central infrastructure questions rather than niche security concerns.
Tools for Humanity, the company behind World ID, has presented the system as a response to this growing need. In Sada’s view, the challenge ahead is to make online spaces trustworthy enough to function in an era where synthetic content and automated interactions are increasingly difficult to separate from human ones.
Whether World ID becomes a standard part of that solution remains to be seen. But the interview makes clear that the debate over how to prove personhood online is moving from theory to an urgent practical problem.