For much of the internet’s short history, we have operated on a quiet assumption that behind every account, post, click or transaction, there is a real human being. But, in a digital world that is now fast being dictated by AI, that assumption is no longer safe.
Nothing characterises this more than the widely-discussed arrival of Moltbook and OpenClaw (formerly known as Clawdbot and Moltbot) – an AI-bot powered social media network and personal AI agent that are demonstrating unprecedented advances in how the technology can function.
Their newfound popularity points to something larger that people are noticing in real time – the growing presence of AI-driven accounts that demonstrate convincingly human characteristics. These are not obviously fake or malicious, they are just present and realistic – demonstrating how the internet and the way we interact online is changing.
AI systems are no longer just tools behind the scenes
Increasingly, AI systems are now looking and acting like real people. In the case of Moltbook, the AI-generated accounts can shift personas with ease. Tone, opinions, writing style, even apparent background can change quickly, sometimes within the same conversation. The conversations are eerily similar to those with human moderators on other social platforms.
What stands out is their flexibility. These systems do not have a single lived identity, and they can adapt faster than any human. To most people scrolling a feed, a Moltbook-style account looks thoughtful, engaged, and emotionally aware. There is often no obvious sign that it is not a person.
When it comes to OpenClaw, these AI-powered bots are actively participating in online discussions. They reply, debate, agree, disagree, and amplify content at scale. Unlike early bots that were repetitive or spammy, these systems can follow context and respond with nuance. They blend into conversations naturally, sometimes engaging dozens or hundreds of people simultaneously.
The key difference here is not intelligence but rather presence. Moltbook and OpenClaw are showing that the boundary and distinction between human and machine participation online is fading fast.
Identity checks are becoming vulnerable
In some cases, AI-assisted actors, because of their increased presence and nuance, are progressively able to pass traditional Know Your Customer (KYC) processes and the most advanced captchas used by banks, fintechs, and other financial institutions.
This is not because institutions and platforms are careless. It is because many KYC systems were designed for a world where identity fraud meant fake documents or stolen passwords, not synthetic personas operating at scale. When AI can generate convincing documents, photos, video, and behaviour patterns, checks that focus only on paperwork, or digitally submitted documents, quickly start to strain.
Importantly, this is not an alarm about widespread failure or collapse, as most institutions are adapting quickly. But it is a reminder that identity verification alone does not always answer a more basic question: is there a real, unique human here?
Human presence now needs to be verified, not assumed
This is where proof of human technology can enter the picture – not as a replacement for KYC, but as a complement to the existing system. Proof of human is a way for someone to verify that they are a real, unique person without revealing who they are. You can prove you are human without sharing your name, financial history, or personal data – you just show that you exist once. This is a different question than identity, as it answers presence rather than profile.
In practice, proof of human gives platforms and institutions an additional signal. Instead of relying entirely on documents or behavioral analysis, systems can ask a simpler question upfront – does this interaction involve a verified human? That proof can reduce fraud, limit abuse, and support fair access without increasing surveillance or forcing people to give up anonymity. Proof of human does not stop AI – it restores balance by giving humans a way to anchor themselves in digital spaces.
A new way to prove we’re human can safeguard trust
This matters because financial systems, like social systems, have always relied on fairness and trust. Whether it’s limits, access, compliance rules or protections, these have all ultimately been designed around the idea that one person equals one participant. When that breaks down, even unintentionally, systems become harder to manage and less equitable.
Moltbook and OpenClaw have both shown what happens when there’s a grey area – conversations become harder to interpret, influence becomes easier to scale and trust becomes fragile. Now, in a world where AI can convincingly participate everywhere humans do, the internet needs a new way to prove humanness without sacrificing privacy. This is not a reason for fear, but rather an opportunity to build systems that work better for humans in the age of AI.
