Emergent, the Indian startup that raised more than $100 million on the premise that millions of non-engineers could build software by describing what they wanted in plain language, has a conversion problem.
Eight million people have used Emergent's vibe-coding tool to create and deploy software. But only 1.5 million of them — roughly 19 percent — are actively using Wingman, the paid AI agent the company launched on April 15 to handle email, scheduling, customer support, and research tasks across Gmail, Slack, WhatsApp, Telegram, and iMessage. Emergent's flagship product has been available for less than 48 hours. The gap between the user base it raised on and the user base it is actually converting is already the most interesting number in the room.
The company is framing Wingman as a safer, more curated alternative to OpenClaw, the open-source AI agent framework that has accumulated 358,000 GitHub stars and a documented security track record that Cisco called "an absolute nightmare from a security perspective." Where OpenClaw allows third-party tools broad access to a system, Wingman requires approval before taking consequential actions — sending mass messages, modifying data — a design choice Emergent calls "trust boundaries." One co-founder told TechCrunch the company spent significant engineering time on security because "we want to be really, really secure."
But the trust-boundary feature is CEO-sourced. No independent security audit has been published. Emergent's claims about Wingman's safety advantage rest on the company's own description of its design choices, not on external verification.
OpenClaw's security posture is, by contrast, documented in detail. Cisco called it "an absolute nightmare from a security perspective," citing the framework's default-open permission model and lack of input sanitization as systemic risks, according to ZDNet. The most critical vulnerability, CVE-2026-25253, carries a CVSS score of 8.8 out of 10 and allows an attacker to take complete control of the gateway and run arbitrary commands, according to Kaspersky threat research. Infostealer malware families including RedLine and Lumma have already added OpenClaw secret file paths to their steal lists.
The comparison Emergent is drawing — paid, audited-by-default versus free, configure-it-yourself — is a real market wedge. Whether it is a true wedge depends on whether Emergent can prove its own house is in order. Its March revenue of $8.4 million represents company-wide earnings, not Wingman-specific revenue, which the company has not broken out. Its $50 million ARR claim from January, targeting $100 million by April, is a self-reported target that has not been independently verified.
The deeper question is whether vibe-coding users — people who write software by describing what they want in plain English — become the same people who pay $20 or $200 per month for an AI agent that reads their email and manages their calendar. Emergent built its $300 million valuation on the first proposition. Wingman is testing the second.
What to watch: whether Emergent publishes Wingman's own revenue or user numbers before the next funding round. If the answer is no, the 81 percent gap stays where it is — a number that explains a lot about why the pivot felt urgent.
