KeyID.ai has launched a free infrastructure platform that gives AI agents real email accounts, phone and SMS access, and website verification capabilities at scale. Announced via a Show HN post by developer vasilyt, agents provision a distinct email identity through a single API call using an Ed25519 keypair — no human setup, dashboard, or API keys required. The free tier supports up to 1,000 separate agent identities simultaneously, a number the company says is sustainable because it runs on a shared rotating domain pool rather than traditional per-mailbox economics.
Behind that model is infrastructure KeyID operates entirely internally: DKIM, SPF, and DMARC configuration, domain warm-up, and reputation scoring are all handled server-side. When an outbound sending domain degrades in reputation, it rotates out of the pool — though the company notes this doesn't affect inbound delivery, so agents keep their assigned address for receiving mail regardless. Rate limiting applies per IP and per identity, and agents triggering spam complaints face throttling or suspension. KeyID ships an MCP server exposing 47 tools via a hosted endpoint at keyid.ai/mcp, alongside Node.js and Python SDKs and a REST API. Supported integrations include Claude, Cursor, CrewAI, AutoGen, LangChain, Playwright, n8n, and Devin. The codebase is open source on GitHub under KeyID-AI/KeyID.
What's drawn scrutiny isn't the infrastructure — it's the use cases KeyID openly markets. The homepage advertises "Registration waves" (agents registering accounts on marketplaces, forums, and partner systems at scale), "Outbound fleets" sending bulk email from thousands of distinct agent identities, and workflows targeting marketplace sellers, social media agents, and lead generation. Nearly every major internet platform's Terms of Service prohibits automated bulk account creation and coordinated inauthentic behavior. Several advertised workflows also collide with CAN-SPAM, CASL, GDPR, and potentially the Computer Fraud and Abuse Act — particularly outbound email operations run from rotating shared domains, which make consistent sender identification and unsubscribe honoring structurally difficult.
KeyID competes against paid alternatives including AgentMail, <a href="/news/2026-03-14-klaus-managed-ai-assistant-hosting-openclaw-pre-bundled-integrations">Klaus</a>, and Instantly.ai, pitching its agent-native design against human-seat pricing models. Pricing beyond 1,000 accounts hasn't been detailed publicly; the company offers a private infrastructure tier for teams needing dedicated domains or isolated sending reputation. As of March 2026, no regulatory or legal action against the platform has surfaced. But zero-attribution identity provisioning paired with explicit fleet-scale account creation marketing puts KeyID in different legal territory than a standard email API — and the company hasn't publicly addressed how it intends to stay there.