Shoppers hoping to snag a quick deal on Amazon were met with frustration instead after a sudden outage left thousands unable to complete purchases, log into their accounts, or even see product prices. Reports of the disruption quickly surged across outage-tracking site Downdetector, with tens of thousands of users flagging problems ranging from checkout failures to glitchy product pages and app crashes. But what exactly triggered the widespread errors, and how long were customers locked out of the world’s biggest online marketplace? Here’s what we know so far about the outage that briefly brought Amazon shopping to a halt.
Amazon responds after surge of outage reports across the U.S.
Digital chaos erupted on the afternoon of March 5, 2026, as Amazon’s vast infrastructure buckled under a wave of technical failures that left tens of thousands of users in a lurch. Shoppers navigating the site encountered more than just broken links; the disruption triggered a surreal experience where product prices fluctuated wildly or vanished entirely, while the checkout process — usually a seamless click — became a wall of payment errors.
Data from Downdetector paints a picture of an unstable recovery effort. After an initial surge of roughly 18,000 complaints at 2:30 p.m. ET, the numbers briefly dipped to 16,000 before roaring back to a new peak. By 3:45 p.m. ET, reports of outages were still hovering near the 20,000 mark, suggesting Amazon was struggling to regain its footing.
In response to the mounting frustration, the retail giant issued a formal apology to USA TODAY as its engineering teams worked behind the scenes to patch the system.
“We’re sorry that some customers may be experiencing issues while shopping. We appreciate customers’ patience as we work to resolve the issue,” Amazon said.
What caused the Amazon outage?
While Amazon has not revealed the root cause of the recent site disruption, the outage appears to be the result of physical infrastructure damage and subsequent technical instability.
The foundational cause appears to be the physical damage to Amazon Web Services (AWS) data centers in the Middle East. On March 1 and March 2, leading up to the retail site’s crash, Amazon confirmed that facilities in the UAE were directly struck by “objects,” causing fires and structural damage. A facility in Bahrain was also impacted by a nearby strike. Local fire departments had to cut power to buildings to fight the blazes, and fire suppression efforts resulted in significant water damage to servers.
Although the suspected physical damage happened outside the U.S., it reportedly triggered a series of technical cascades that eventually hit the global retail site on March 5.
The damage is believed to have crippled Amazon S3 (storage) and Amazon DynamoDB (database), which are foundational services. Because the global retail site relies on these for product listings and user data, the regional loss of capacity put immense strain on other data center clusters (like US-EAST-1) to pick up the slack.
As previously mentioned, during the peak of the outage, shoppers saw fluctuating prices and vanishing listings. This is a classic symptom of DynamoDB being unable to serve consistent data across different regions, leading to the checkout and payment errors reported by over 20,000 users.
Is Amazon back up and running?
As of this writing at 7:15 p.m. ET, Amazon appears to be largely back up and running, though it is still recovering from the massive disruption earlier today.
While the core checkout and login errors that defined the peak of the outage have been mostly resolved, some users are still reporting minor aftershock glitches, such as slow loading times or issues with order tracking.
Recent user comments on Downdetector indicate that checkout processes have been functioning normally as of late.
Sources: USA TODAY, Downdetector