Filing for a domain extension and actually running one turn out to be very different businesses, and Aruba just bet on 39 of them at once.
The European hosting provider submitted applications for 39 new generic top level domains during ICANN’s 2026 expansion round, a list that ranges from the obviously strategic to the genuinely surprising. .vps sits closest to home, given that Aruba already sells virtual private servers, cloud hosting, and email under one roof. .lab and .demo could appeal to developers and testing environments. Then things widen out fast, into .gaming, .bnb, .gym, .zen, and even the geographic .roma, territory that has little to do with the infrastructure business Aruba actually runs.
What makes this more than a land grab is that Aruba has done it before. The company already operates .cloud, secured during the last major expansion round, and that experience covers real operational weight: DNS infrastructure, registrar relationships, abuse handling, and the kind of continuous uptime regulators and customers expect from a registry. Filing paperwork does not require any of that. Running a registry for years does.
That distinction matters because ICANN’s window closed in August with more than 1,600 applications on file, and plenty of the popular strings will end up contested by multiple companies before any of them reach the internet’s root. Some of Aruba’s 39 will likely get whittled down through evaluation, and others could stall out completely.
Even the ones that survive face a second, quieter problem. A registry only works if registrars want to sell it and customers want to renew it once the introductory discount wears off. The last expansion round left behind a long list of extensions that pulled in cheap signups and then watched renewal rates collapse. Aruba spreading its bets across infrastructure names, trust services like .pec and .sign, and consumer terms looks like an attempt to avoid repeating that mistake by not putting everything on one theme.
Owning .cloud proved Aruba can run a registry. Thirty nine applications are testing something harder: whether that know-how scales into a real business, or just a long list of domains nobody particularly needed.
