Amazon Web Services unable to restore Bahrain data network knocked out by Iran

What happened

Amazon Web Services has confirmed it will not be able to restore one of its three UAE data hosting zones or its Bahrain network, both of which were damaged amid the Iran conflict that began on 28 February. In an update posted to its Health Dashboard, AWS said "thorough assessments" led it to conclude it is unable to restore access to resources and data hosted exclusively in those locations, and that it has notified authorities in both Gulf states.

The disruption to the affected UAE zone began in March. AWS said most customers in that zone have since re-established operations elsewhere by restoring backups or copying data that remained accessible, and that customer support teams remain available to help remaining users migrate to alternate regions. AWS operates nine sites across the Middle East, three each in the UAE, Bahrain and a third location.

Why it matters

This is a rare admission from a hyperscale cloud provider that a regional zone cannot be recovered at all, rather than restored after an extended outage. For enterprises and public-sector bodies that built their business-continuity plans around the assumption that a named AWS "Availability Zone" is effectively indestructible, this is a direct test of that assumption in a live regional market.

For digital transformation leaders in the Gulf, the episode sharpens an existing question about data residency, redundancy and sovereign resilience: how much critical infrastructure should sit in a single physical zone, however well-architected, when geopolitical risk is part of the operating environment. It also puts a spotlight on how clearly cloud providers communicate irreversible loss, versus temporary disruption, to customers who may still believe recovery is pending.

By the numbers

• Nine AWS sites currently operate across the Middle East, three each in the UAE, Bahrain and a third hub.
• Three UAE data hosting zones exist in total; one is confirmed unrecoverable.
• The disruption to the affected UAE zone began in March, following the conflict's onset on 28 February.

The Renascence take

The headline fact here isn't the attack itself — it's the communication moment. AWS has moved from "disrupted" to "permanently unrecoverable," and that shift changes what good customer experience looks like: the job is no longer reassurance, it's closure, clarity and active migration support.

Most customers don't fail at backups — they fail at believing the backup plan will never actually be invoked. The real lesson for any organisation relying on a single zone, cloud or otherwise, is that resilience has to be designed, tested and rehearsed before the loss notice ever arrives, not after. Gulf enterprises and regulators now have a concrete, named precedent to point to when building the business case for multi-region or sovereign-redundant architecture — and cloud providers have a model to improve on for how plainly and how early they tell customers when "disrupted" has quietly become "gone".