World Brings Proof of Human to Zoom Calls and Banks as AI Fraud Heads Toward $40 Billion

Signup Write New Story World Brings Proof of Human to Zoom Calls and Banks as AI Fraud Heads Toward $40 Billion by Ishan Pandey by Ishan Pandey | @ ishanpandey AI, Web3, Cybersecurity, Startup Funding & Enterprise SaaS. Top Journalist, Analyst & Thought Leadership Media Leader.

AI, Web3, Cybersecurity, Startup Funding & Enterprise SaaS. Top Journalist, Analyst & Thought Leadership Media Leader.

Americans reported $20.9 billion of cybercrime losses to the FBI in 2025, six times the $3.5 billion logged in 2019, while Deloitte expects US losses from AI-enabled fraud to climb from $12.3 billion in 2023 to $40 billion by 2027. Against that backdrop Tools for Humanity joined the Zoom ISV Exchange Program on October 19 so that World ID for Zoom is now sold through the Zoom App Marketplace to the platform's business customers. World co-founder Alex Blania unveiled World ID for banking and finance at Money20/20 as the first industry-specific package built on the same Proof of Human technology.

The FBI's Internet Crime Complaint Center recorded $3.5 billion of losses in 2019 and $12.5 billion in 2023, after which the 2024 report put the figure at $16.6 billion and the 2025 report reached $20.9 billion across 1,008,597 complaints, a 26% rise in a single year with investment fraud alone accounting for more than $8.6 billion of it. A line that climbs every year for six years in a row describes a market in which each new tool for criminals finds buyers faster than banks can adapt their defences, which is why financial institutions are now looking at how to confirm identity at the moment of contact on top of how strong any one password is.

The Deloitte Center for Financial Services projects that fraud losses enabled by generative AI in the United States will grow from $12.3 billion in 2023 to $40 billion in 2027, a compound annual rate of 32% that Deloitte ties to cheap fraud software sold on the dark web for anywhere between $20 and several thousand dollars. The same research expects business email compromise, which cost Americans about $2.7 billion in 2022, to turn into $11.5 billion of generative-AI email fraud by 2027 under an aggressive adoption scenario, which tells banks that the cost of a convincing fake is falling toward zero while the value of a successful one keeps rising.

The case that made the threat concrete for treasurers was the one at engineering group Arup, where a finance employee in Hong Kong joined a video call with people who looked and sounded like the company's chief financial officer and colleagues and went on to send HK$200 million, about $25.6 million, across 15 transfers after police later determined that every other participant was a deepfake re-creation. Arup's chief information officer Rob Greig said attacks of this kind had been rising sharply in the months around the incident. The data since then backs him up, because Pindrop measured a 1,300% jump in deepfake fraud across more than 1.2 billion customer calls in 2024 with US incidents rising from about one a month to seven a day and synthetic-voice attacks up 149% at banks, 475% at insurers and 107% at retailers while face-swap attacks grew 704%.

The FBI's 2025 report is the first to carry an AI descriptor on complaints. The descriptor already attaches to $893 million of losses across 22,364 complaints with $632 million of that sitting in investment fraud and a further $30 million in business email compromise. Because the tag applies only when a victim says AI was involved, the bureau itself treats the figure as an undercount. Since investment fraud is the category in which banks, brokers and asset managers carry the most customer exposure, the pattern points straight at the financial sector as the place where proof of who is on the other end of a call will earn its keep first.

Membership in the Zoom ISV Exchange Program allows Zoom to offer selected independent software vendor products directly to customers through its own platform, so World ID for Zoom can now be purchased inside the Zoom App Marketplace by any business that wants to verify that meeting participants are real people. Those businesses can also request Orbs, the hardware used for one-time Proof of Human verification, for delivery to their offices. The channel is large, since Zoom last disclosed about 192,600 Enterprise customers, reported 4,468 customers paying more than $100,000 a year at the close of fiscal 2026 on revenue of $4.87 billion and lists more than 3,100 apps in a Marketplace that now carries a product aimed squarely at the deepfake problem.

The integration runs on Zoom's Realtime Media Streams and is hosted on AWS. It confirms presence through a three-way match between the cryptographically signed image captured when the participant first verified at an Orb, a liveness selfie taken in real time on the participant's own device and the live video frame that the other participants see on screen. When all three agree the participant's video tile shows a Human Badge, which means a host looking at a chief financial officer on screen now has a hardware-backed answer to the question of whether the person who enrolled at an Orb is the person talking. The other party can complete the face-authentication steps in a few moments instead of waiting for software to guess whether pixels look synthetic.

The new suite for banks combines the World ID app, an identity wallet in which people store and present the Proofs that show they are human, with World ID for Zoom for live communication and World ID for Agents, which lets a person establish Proofs about themselves and delegate them to software that then confirms those Proofs to a third-party platform. World is working with design partners OpenExchange, Bain Capital, VanEck Funds and Outtake to test the suite across customer accounts, employee workflows and AI agents. The launch marks the first time World ID has been packaged for a single industry, a choice that reflects how many consequential decisions in finance (trades, transfers, onboarding and approvals) still rest on someone being who they say they are.

Tools for Humanity announced World ID partnerships with Zoom, Docusign, Okta, Vercel and Tinder at its April event in San Francisco when about 17.9 million people had signed up, Zoom opened an enterprise beta in May that VanEck joined. The general release arrived in October together with the banking package, so a product that did not exist in public form in the spring is now on sale to regulated institutions in the autumn. More than 18 million people now hold a Proof of Human. The suite pairs that base with cryptographic Proof verification that World describes as quantum-secure.

Software that acts for people is the third leg of the suite. The numbers around it explain the urgency. Coverage of AgentKit reports that agents registered to the ERC-8004 identity standard rose from a few hundred at the start of 2026 to about 130,000 by June, that the x402 payment rail built with Coinbase and Cloudflare has cleared more than 100 million agent transactions, that Cloudflare sees more than 10 billion AI bot requests a week and expects bot traffic to pass human traffic on its network by 2027 and that McKinsey projects agentic commerce at $3 trillion to $5 trillion by 2030.

A person who delegates a Proof to an agent lets a bank, an exchange or a retailer see that a verified human stands behind a request without learning who that human is, which is the property institutions need when a customer's assistant starts moving money on their behalf.

Ajay Patel, Head of World ID and Chief Revenue Officer at Tools for Humanity, describes Proof of Human as an essential layer of trust infrastructure that enterprises can no longer treat as optional. He sees the Zoom partnership as the way to bring that layer into the workflows where a wrong answer costs the most.

Sandro Herbig, president at Tools for Humanity, says enterprises are now asking people to present Proofs about themselves as a way to reduce fraud and raise trust and that the banking suite gives them one set of tools for internal systems and for customer-facing ones. World was conceived by Sam Altman, Max Novendstern and Alex Blania with the aim of building the largest and most inclusive network of real humans. The banking launch shows how that consumer-scale network is being converted into infrastructure that a compliance department can buy through a channel it already uses.

World will introduce the technology to more enterprise buyers at Zoomtopia and through its activation at Cloudflare Connect. The combination of a distribution channel with hundreds of thousands of business accounts, a design-partner group drawn from asset management and banking plus an identity base above 18 million people gives the company a real route into the institutions that lose the most money to impersonation. If the Deloitte forecast holds and AI-enabled losses reach $40 billion by 2027, the banks that can show a Human Badge on a video call and a delegated Proof behind an agent's request will have an answer ready for regulators and customers alike. World has set itself up to be the supplier they call first.

Don’t forget to like and share the story!

Vested Interest Disclosure: HackerNoon has reviewed the report for quality, but the claims herein belong to the author. #DYOR.

Forgepoint Raises $9.8M From Primary and a16z Speedrun to Give Factories an AI Sales Team

XRPL Commons Hosts 36-Hour NYC Hackathon to Build On-Chain Lending and AI Payments on XRPL