World is expanding World ID, its proof-of-human protocol, to bring human verification into AI agents, enterprise workflows, dating apps, meetings, digital agreements and ticketing systems. The next-generation World ID was announced during Lift Off, a live event in San Francisco, with integrations involving Docusign, Okta, Tinder, Vercel and Zoom.

The rollout comes as AI-generated identities, automated agents, deepfakes and bot activity make it harder for platforms to know whether a real person is behind an online action. World says World ID has already been used more than 450 million times, and the new version expands the protocol into settings where human verification may affect trust, access, fraud prevention and accountability.

For businesses, developers and consumer platforms, the decision is whether proof of human should become part of the trust layer for AI-era workflows. That question is becoming more urgent as agents begin browsing the web, accessing APIs, completing purchases and executing multi-step tasks on behalf of users.

In short, World ID is becoming a verification layer for digital systems that need to know whether a real human is behind an action, profile, meeting, document, agent or transaction. The update connects proof of human to both everyday platforms and higher-stakes workflows where AI-generated activity can create new trust and accountability challenges.

World ID is a proof-of-human protocol that allows a person to privately verify they are a real, unique human across digital services without exposing personal identity information.

World ID is expanding from proof of human verification into a trust layer for AI agents, enterprise workflows and consumer platforms.

World announced the next generation of World ID, describing it as a proof-of-human protocol built to increase trust across digital services as AI makes fake content, identities and interactions more convincing. The announcement was made during Lift Off, a live event in San Francisco.

Tiago Sada, Chief Product Officer at Tools for Humanity, said the trust problem is becoming more urgent as online activity becomes easier to fake. “If anything online can be faked, you no longer know who or what to trust,” Sada said. He added that World ID, which World says has been used more than 450 million times, is meant to provide “a simple, privacy-preserving way to prove there’s a real human behind every interaction.”

The announcement is part of a larger change in how online activity is being carried out. AI agents are already browsing websites, accessing APIs, completing purchases and executing multi-step workflows on behalf of the people who use them. But many of the systems those agents interact with were built for human users, where trust is often tied to accounts, credentials or payment methods rather than verifiable human approval.

That creates a trust gap for businesses and developers. When an agent signs up for a service, accesses an API, makes a purchase or changes a workflow on behalf of a human, platforms may not have a reliable way to confirm that a real human approved the action. World is presenting proof-of-human verification as the missing trust layer: a way for services to verify, using privacy-preserving digital proof rather than personal data, that a real, unique human stands behind an agent and its actions.

The issue becomes more serious when agents handle higher-stakes work. Payments have card networks and payment protocols that can help confirm whether an account holder authorized a transaction, but many agent actions happen outside payment systems. An agent may access sensitive data, sign a contract, modify infrastructure or approve a deployment, and those actions still need a reliable record of human intent.

The risk is easy to understand: an agent could make a large purchase that a user later disputes, or change production infrastructure in the middle of the night without a clear record that an engineer approved it. In both cases, the problem is not only that the agent acted. The problem is that the platform may not be able to prove a human was involved at the moment the decision mattered.

That uncertainty also raises a larger accountability question. As AI agents begin taking actions on behalf of people, businesses and platforms will need clearer records of authorization: who approved the action, whether the agent was acting within delegated permission and who is responsible if the action causes harm, cost or disruption. World’s argument is that proof-of-human verification can help create that record without requiring every platform to collect more personal data.

The release brings World ID into several everyday and business-critical environments, including Docusign, Okta, Tinder, Vercel and Zoom. The common thread across the integrations is the same: World wants platforms to verify that a real human is behind an action, account, agent or interaction without requiring that person to reveal their identity across every service they use.

World’s human-in-the-loop capability, powered by World ID, is designed to address that gap. It allows an agent workflow to request privacy-preserving digital proof that a unique real human authorized a specific action, with the proof remaining independently verifiable later. The system uses a zero-knowledge approach, meaning a platform can confirm that a real human approved the action without seeing or storing the person’s private identifying information. World says that reduces personally identifiable information risk and limits the chance that verification data becomes another source of leakage.

World is also working with Vercel to bring human-in-the-loop verification to developers building on Vercel’s open-source Workflow SDK. Developers can add a human verification step to a workflow or agent, and every verification is visible inside the workflow execution for auditability. World says the feature is live today and can be installed from npm.

World describes the new World ID as a full-stack upgrade for proof-of-human verification, built around an account-based architecture that allows verification to carry across apps, services and devices. World says the new architecture includes several changes meant to make proof-of-human verification more usable in real-world settings:

The key point: World is trying to make proof-of-human verification reusable across apps, services and devices without turning it into a traditional identity system. That matters because platforms need stronger assurance that a real person is involved, while users and businesses still need privacy protections that limit unnecessary data collection.

World is also extending World ID into enterprise environments where the cost of impersonation, fraud or AI-generated deception can be high.

Zoom is integrating Deep Face, World’s face-verification technology, directly into its meetings product, making it the first communications platform to bring World ID-based human verification into live video meetings. The integration is built around World’s Orb, the physical device World uses to verify that a person is a real, unique human. When a participant first verifies with an Orb, the system captures a signed reference image that can later be used as part of a privacy-preserving verification check.

For Zoom meetings, World says the system uses a three-way match. It compares the signed image from the participant’s original Orb verification, a real-time Face Auth liveness selfie taken on the participant’s device to confirm the person is physically present, and the live video image that other participants see in the Zoom meeting. In plain terms, the system is checking whether the person who verified with World, the person confirming their face on their device, and the person appearing on Zoom all match.

When those three checks match, the integration is meant to confirm with high assurance that the person appearing in the Zoom meeting is the real, verified human expected to be there. World says the integration analyzes video only, not audio.

For business meetings, the concern is not only whether someone has the right login credentials or meeting access. As deepfake tools improve, organizations face a growing risk of video-based impersonation and fraud, especially in meetings involving approvals, hiring, financial decisions or sensitive company information. World’s Zoom integration is aimed at that video identity problem: confirming that the visible participant matches a previously verified human.

For digital agreements, Docusign and World are working to bring proof-of-human verification into the document signing trust model. Through World ID, signers can confirm specific attributes about themselves, including that they are human and not a bot. World says this creates a foundation for human continuity in agreement workflows, giving document actions a clearer tie to a verified human whether the action is performed directly by a person or delegated to an agent.

That matters for business workflows because AI agents may increasingly prepare, route or trigger actions inside agreement processes. World’s point is that even when software handles part of the workflow, trust still depends on linking important actions back to a verified human.

World is also working with Okta on how digital services verify the relationship between humans and agents. As AI agents become more common, developer products with free offerings and social media platforms face a growing risk of scaled abuse, fraud and automated misuse. Without a way to distinguish the human behind an agent, services may have to rely on aggressive rate limits or heavier Know Your Customer processes, which can add friction, introduce privacy risks and reduce the value of automated agentic workflows.

World ID allows a verified human to delegate proof-of-human verification to an agent so it can act on their behalf. Services receiving requests from that agent can confirm the activity is tied to a real, unique human without collecting personal data.

To address this, Okta is planning a new product called Human Principal, which will allow API builders to verify whether a human stands behind an agent and its actions, then enforce policies based on that verification. Under Okta’s planned approach, humans will be able to verify themselves through several verification methods and obtain device-bound digital proof that can carry across products without repeated verification.

World says World ID is slated to become one of the first Human Principal integration partners, providing a privacy-preserving proof-of-human method. Once combined, World ID and Human Principal will enable features such as rate limits per human for agent traffic, abuse-protected free tiers and a cleaner onboarding flow for agents that need to access services on behalf of the person they represent.

The waitlist for the upcoming Human Principal beta is open at humanprincipal.ai.

World is also expanding proof-of-human verification into agentic workflows through AgentKit, which includes three capabilities: agent delegation, human in the loop and agentic commerce. Together, those tools are meant to help developers build AI agents that can carry proof of the human behind them, request human approval for sensitive actions and support commerce actions on behalf of verified humans.

In commerce, the stakes are immediate because bots already exploit high-demand moments such as flash sales, limited inventory, exclusive product releases and promotional pricing. Shops may want to support AI agents acting on behalf of real customers, but they also need a way to distinguish legitimate agent traffic from unwanted automated activity.

With AgentKit, a buyer can delegate their World ID to an agent so the agent can find products online, claim coupons and help with the purchase process. Human-in-the-loop verification can require explicit approval before an agent completes purchases above a set threshold. That gives merchants a way to confirm one verified human per allocation and limit how many units each buyer can purchase.

World says these capabilities are intended to prioritize legitimate buyers, support revenue opportunities and reduce fraud. In a demo using the Universal Commerce Protocol, an open standard for agentic commerce co-developed by Shopify and Google, World showed how Shopify merchants could verify that a purchase is backed by a unique, verified human: one person, one agent, one allocation.

Developers interested in the agentic commerce skill can explore World’s AgentKit Shopify demo on GitHub.

As AI agents begin evaluating tools, accessing APIs and acting on behalf of buyers, companies face a practical tension: they want to give agents useful access without opening the door to abuse. World says World ID helps address that problem by allowing services to distinguish verified human-backed agents from unverified automated traffic.

Browserbase, which builds a platform for agents to automate web activity, has integrated agent delegation. When an agent using Browserbase carries a World ID, World says it can receive the benefits of verified traffic, reducing blocks and friction as it navigates the web. Without that verification, the agent encounters standard anti-bot protections. The integration is live for developers.

Exa, which builds AI-native search, is offering agents verified through AgentKit 100 free API requests per month. World says verified agents backed by World ID can experiment with Exa’s search and content endpoints without upfront payment, while Exa remains protected against abuse. After the quota is exhausted, the agent falls back to standard x402 payments, a pay-per-use system that lets AI agents automatically pay for API access instead of relying on a traditional subscription or account setup. The integration is also live for developers.

World is also expanding World ID into consumer platforms where authenticity, fairness and access can affect real-world costs for everyday users.

On Tinder, World and Match Group are expanding a World ID pilot that first launched in Japan last year. The integration is expanding in select markets, including Japan and the United States. Users can verify with an Orb once, connect their Tinder account and display a verified human badge intended to show that a real person is behind the profile. World says the badge can help encourage higher-quality connections by showing that users are real. For a limited time, verified users can also receive five free Boosts on Tinder, a paid in-app feature that makes a profile one of the top profiles in a local area.

Live events are another consumer trust use case because bots can drive up costs and limit access for real fans. World introduced Concert Kit, a product that allows artists to reserve tickets for verified humans. Artists are already committing to piloting Concert Kit. World says Thirty Seconds to Mars will reserve a portion of tickets for verified humans on its upcoming 2027 tour, with more details to come.

World also addressed reports suggesting that Tools for Humanity had a partnership or commercial relationship with Bruno Mars, The Romantic Tour or Live Nation. World says that is not the case, and that there is no business relationship, agreement or collaboration between Tools for Humanity, World or related products and those parties.

Q: What did World announce with the new World ID?
A: World announced the next generation of World ID, its proof-of-human protocol, with new architecture and integrations across Docusign, Okta, Tinder, Vercel and Zoom. The update expands human verification into AI agents, enterprise workflows, consumer platforms, digital agreements, meetings and ticketing systems.

Q: How does World ID verify a human behind an AI agent?
A: World ID uses AgentKit capabilities such as agent delegation, human in the loop and agentic commerce. These tools allow an AI agent to carry proof of the verified human behind it, request verifiable approval for sensitive actions and support transactions where a platform needs to know that one real human stands behind the agent.

Q: Why does World ID matter for AI agents now?
A: AI agents are beginning to browse websites, access APIs, make purchases and complete multi-step tasks for users. World ID gives platforms a way to verify human involvement in those actions without requiring every service to collect more personal data or create its own heavy identity process.

Q: What parts of the World ID rollout are live now?
A: World says Vercel’s human-in-the-loop integration, Browserbase agent delegation and Exa’s verified-agent API access are live for developers. Agent delegation, human in the loop and agentic commerce are also available through AgentKit.

Q: What parts of the World ID rollout are still planned or limited?
A: Okta’s Human Principal is planned and has a beta waitlist, while Tinder’s World ID integration is expanding in select markets, including Japan and the United States. World also says future products, features, functionalities or certifications mentioned in its announcement are informational and should not be treated as commitments to deliver.

Q: How is World ID being used outside AI agent workflows?
A: Zoom is using World ID with Deep Face for meeting identity assurance, Docusign is working with World on proof of human for agreement workflows, Tinder is expanding verified human badges, and Concert Kit will allow artists to reserve tickets for verified humans.

World’s announcement points to a practical trust problem for the AI era: platforms need to know when a real human is behind an account, action, agent or transaction.

Key point: World ID is trying to make online actions more trustworthy by showing when a real human is behind them. As AI agents take on more tasks, platforms need a reliable way to know whether a person approved an action, delegated it to an agent or can be connected to the activity when accountability matters.

Who should care: Developers, enterprise platforms, marketplaces, collaboration tools, commerce companies and consumer apps should pay attention. Products or services that allow agents, bots, profiles or automated workflows to interact with users may need better ways to separate human-backed activity from unverified automation, especially when trust, access, identity or authorization is at stake.

Why this matters now: AI agents are starting to browse, transact, access APIs and complete work across systems that were originally built for human users. As those agents move from low-risk tasks into higher-stakes actions, companies need human verification that can confirm approval or accountability without requiring unnecessary personal data collection.

What decision this affects: Businesses will need to decide whether proof of human verification belongs in their trust, fraud prevention, agent access or identity strategy. Developers building agent workflows will also need to decide when human verification should be required, especially for actions involving money, contracts, infrastructure, personal data or limited inventory.

In short, World ID is trying to answer a question every digital platform will face as AI agents become more capable: who is the real human behind the action? The answer will matter not only for security, but also for privacy, access, fairness and accountability across AI-era services.

The next test of online trust may be whether platforms can prove the one thing that matters most: a real human was behind the decision when it counted.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing support, AEO/GEO/SEO optimization, image concept development, and editorial structuring support from ChatGPT, an AI assistant. All final editorial decisions, perspectives, and publishing choices were made by Alicia Shapiro.