Summary

  • In two 2026 RIPE Labs essays, Hisham Ibrahim asked readers to map the hidden dependencies behind digital-sovereignty plans: suppliers can differ on paper while sharing an upstream, software stack, legal jurisdiction or cloud region.
  • His commons test asks whether a measure gives people usable choices and keeps networks interoperable. Procurement, local infrastructure and crisis powers produce different answers depending on how they are designed and whether people can change course.

A spare provider looks excellent in a tender. It can fail as a spare if both providers rely on the same upstream. This is a hypothetical, but the mechanism comes straight from Ibrahim's June RIPE Labs essay. He listed the hidden common parts: a control plane, a software stack, a jurisdiction, a cloud region, even an operating assumption. The invoice shows vendor names. The failure sees the shared component.

Ibrahim calls the map of things that must continue working dependency topology. That phrase earns its keep. A network operator can draw one path from an application to a cloud region, another through DNS and transit, then trace where the lines meet. A government buying two services can ask the same question. How much of the apparent choice survives the loss of a common supplier or a policy change at a common control point? His test moves the conversation from flags on servers to the ability to keep service running.

The person making this argument is RIPE NCC's Chief Community Officer, whose remit includes member engagement, learning and external coordination. That helps explain why his essay treats operator skill and local relationships as infrastructure. An additional cable or exchange point is useful when people can operate it, troubleshoot it and connect it to viable alternatives. A procurement clause has no hands to move traffic during an outage.

His July essay turns that observation into a commons test. A public buyer can demand open standards, local expertise and a genuine choice of providers. It can also anoint one mandatory platform and call the resulting dependence national capability. The same instrument—public procurement—can widen or narrow the exits. Regional investment faces a similar question: does it improve interconnection and global reachability, or leave a tidy local island with too few paths out?

Ibrahim also draws a line around the Internet's shared coordination functions. Registry identifiers and open protocols make independent networks work together. Cloud contracts, AI capacity and public digital services sit on other layers. At a September RIPE NCC session with the ITU and Lebanon's mission, he made that distinction between technical coordination, Internet governance and broader digital governance explicit. A policy aimed at a cloud dependency needs its sights on the cloud dependency. Moving the lever to the common naming or numbering layer can put every network's interoperability in the path of a narrower dispute.

The best line in his argument is a practical question, not a slogan: can the affected people understand the dependency, choose another path, and recover when the first one fails? If the answer depends on the same supplier operating both exits, a sovereignty plan has bought two doors to one room.