OpenComputer dot dev presents itself as a persistent cloud VM service built specifically for AI agents. The idea is straightforward: instead of disposable sandboxes, you get a lasting computer that hibernates and wakes on demand. On the surface, the site looks polished, with an About page, privacy policy, terms of service, and even a German Impressum — more legal structure than many new SaaS tools bother with.
But here's where the caution comes in. For a company asking you to run code on its infrastructure, the ownership is completely opaque. The domain registration is hidden behind a privacy wall, and we could not find the actual people or registered business behind it. That matters because when something goes wrong — a data breach, a billing dispute, a disappearing service — you need to know who to hold accountable. Legitimate infrastructure providers like DigitalOcean or Linode have clear corporate identities, public leadership, and years of operational history. OpenComputer has none of that.
The site also has no past to examine. There are no Wayback Machine archives, no Trustpilot reviews, and no traffic ranking. It may be brand new, but for a service that processes potentially sensitive agent workflows, the absence of any third-party track record is a real unknown. The technical setup is competent but not polished: modern encryption is in place, but basic browser security protections are missing, and old insecure protocols are still accepted.
So is opencomputer dot dev a scam? There's not enough evidence to call it that. But is it safe to trust with your workloads and payment details? Right now, the answer is no. Until the company puts real names and a verifiable business address behind the product, treat this like any other unknown infrastructure provider: test with throwaway accounts and non-sensitive data, and don't commit anything you can't afford to lose.