Microsoft confirms rerouting delays after subsea cable faults; experts warn of rising costs, fragile connectivity, and urgent need for resilience planning.
Undersea cable damage in the Red Sea disrupted internet access across parts of Asia and the Middle East over the weekend and caused latency issues for Microsoft Azure services routed through the Middle East. Microsoft confirmed the incident, noting that while network traffic was not interrupted, rerouting through alternate paths introduced delays.
The latency issues began at 05:45 UTC on 06 September 2025. In its latest update at 19:52 UTC on 07 September, the Microsoft Azure status page stated, “We do expect higher latency on some traffic that previously traversed through the Middle East. Network traffic that does not traverse through the Middle East is not impacted. We’ll continue to provide daily updates, or sooner if conditions change.”
Even Akamai Cloud’s Linode Status reported emerging connectivity issues. “Linodes in our IN-MAA, AP-West, and IN-BOM data centers are experiencing increased network congestion and latency due to multiple faults in the undersea cables that are part of the optimal routes out of the data centers,” the Linode Status page stated.
Internet monitoring group NetBlocks reported widespread connectivity degradation across multiple countries in the region, linking the disruption to failures on subsea cable systems in the Red Sea.
Microsoft, Linode, and NetBlocks did not immediately respond to a request for comment.
High latency for enterprise apps
For CIOs and enterprises, the Red Sea cable disruption is more than a technical hiccup. It directly translates into slower applications, higher operational costs, and potential business risks.
“The recent subsea cable cuts have forced traffic onto longer alternative routes, causing noticeable latency spikes for cloud-native applications sensitive to round-trip times, including real-time APIs, trading platforms, video conferencing, and ERP systems. Enterprises may face slower web apps, delayed database queries, longer file transfers, and lag in multi-region replication,” said Manish Rawat, analyst at TechInsights. He added operationally, cloud-dependent businesses, including logistics, e-commerce, and SaaS, could see temporary service disruptions affecting revenue, SLAs, and customer experience.
Such disruptions highlight the fragile dependency of global enterprises on undersea cables, where a single break in a chokepoint region can ripple into delayed services and missed business commitments worldwide.
“Latency and rerouting raise costs for both sides, providers and enterprises. Providers face higher OPEX from buying alternate transit, managing traffic overhead, and adding temporary long-route capacity, while enterprises absorb indirect costs from slower applications, lost productivity, reputational damage, and transaction losses,” said Pareekh Jain, CEO at EIIRTrend & Pareekh Consulting.
Beyond the immediate latency costs, outages of this scale trigger a chain reaction of hidden financial costs as well.
“The cost of repair is steep and slow: millions of dollars and weeks of downtime for each cable,” said Sanchit Vir Gogia, chief analyst and CEO at Greyhound Research. “Hidden costs pile up as well. Insurers raise premiums on infrastructure in conflict-prone waters. Regulators question the adequacy of resilience planning.”
In addition, reputational damage lingers when services appear unreliable.
What went down at the Red Sea
NetBlocks has reported a series of subsea cable outages in the Red Sea that affected the SMW4 and IMEWE cable systems near Jeddah, Saudi Arabia, impacting internet connectivity in the UAE, Pakistan, and India.
The SMW4 (South East Asia-Middle East-West Europe 4, SEA-ME-WE 4) is an approximately 18,800 km submarine cable that was ready for service on December 13, 2005. Consisting of two fibre pairs, the submarine cable connects 14 countries, including Singapore, India, Pakistan, the UAE, Saudi Arabia, Egypt, Italy, and France. The 12091km submarine cable IMEWE ( India-Middle East-Western Europe) links India & Europe via the Middle East. It consists of three fibre pairs, with 10 landing stations in eight countries.
Reports indicate Yemen’s Houthis rebels are behind the outage, but Houthis have denied attacking the lines in the past. The cause of this specific disruption hasn’t been independently verified yet.
“Globally, over 95–97% of international data traffic travels through underwater fiber-optic cables, not satellites or terrestrial networks. The Red Sea is a critical chokepoint in this ecosystem as an estimated 17% of global internet traffic passes through its underwater cables, with roughly 80% of traffic from Asia to Western countries moving through these routes,” said Amajit Gupta, group CEO & MD at Lightstorm.
He added Red Sea has long been a strategic chokepoint for global connectivity – dense with cable traffic, historically sensitive due to regional instability, in addition to being exposed to geopolitical risks, shipping lane hazards, and natural disasters.
Wake-up call for CIOs
This is not an isolated incident. In February 2024, three undersea cables were damaged in the Red Sea, impacting internet connectivity between parts of Asia, Africa, and Europe. For CIOs, the event only reinforces the need to rethink resilience strategies. Experts highlight that relying solely on hyperscalers’ rerouting is no longer enough. Diversifying routes across regions, adding satellite or terrestrial backups, and maintaining active monitoring systems should be the priority.
“They should review SLAs, build cloud contingency plans, and collaborate with providers on resilience options and traffic prioritization. Regular drills and simulations help validate readiness and adjust processes for outages, latency, or rerouting events,” added Jain.
Greyhound Research’s Gogia said CIOs must demand evidence of true physical diversity, not just logical redundancy, in their provider relationships. “They need clarity on how much traffic flows through the Red Sea, what alternatives exist, and how quickly those alternatives can be activated.”
