Users from all over the world had reported problems with appsAI assistants and video games that depend on Amazon Web Services.
San Francisco/Madrid, October 20 (Xinhua/EuropaPress).- The services in the cloud of Amazon were recovering this Monday after being offline for hours, interrupting the services of some websites y applications popular worldwide.
The problem “has been completely mitigated and most of the operations of the service AWS “operating normally now,” the company’s maintenance website said in an update. Amazon Web Services (AWS)and added that some services They still have effects.
It said its engineers are “making progress addressing the issue” that they first noticed as an “increase in error rates and latencies across multiple AWS services in Eastern US Region 1.”
Amazon Web Services worked to solve an interruption in its services that caused the paralysis, at a global level, of platforms such as Amazon, Alexa, Snapchat, Epic Games Store or Fortnite, in addition to the services of the Artificial Intelligence (AI) assistants ChatGPT from OpenAI and Perplexity.
✅ Service issues are resolved! Epic Games Store and Epic Online Services are fully recovered and have all returned to normal.
We’ll continue to monitor this to ensure everything remains stable.
— EOS Status (@EOSStatus) October 20, 2025
The company reported an increase in error rates and latencies for several AWS services on Monday, initially linked to the US-EAST-1 US region. However, service provider issues were subsequently also identified in several global regions, including Spain.
As a result, users experienced the interruption of various services, both in web pages and in applications and video games, that depend on this infrastructure and that did not operate correctly. This is the case of Amazon itself, with Amazon Prime Video and its assistant Alexa, which did not respond to requests, as some users shared on Reddit; as well as Epic Games Store and Fortnite, as detailed by the video game company in a publication on the X social network.
Likewise, failures due to this incident with AWS were also reported by the artificial intelligence (AI) companies OpenAI – which experienced interruptions in ChatGPT – and Perplexity, which made direct reference to the problem with Amazon in another publication on X. Other users reported problems with platforms such as Snapchat, Duolingo or Canva.
Specifically, as Amazon specified in a statement on its AWS status check page, the cloud infrastructure service interruption appeared to be related to the DNS resolution of the DynamoDB API endpoint in the United States (US-EAST-1), where the failures began to be reported.
However, the company clarified that it was working on several parallel routes to accelerate the recovery of its services and, therefore, of the companies that use its infrastructure. “During this time, customers may not be able to create or update support cases. We encourage customers to retry any failed requests,” Amazon added.
With all this, in its most recent update on the case, Amazon had announced that mitigation measures were applied and that “significant signs of recovery” were being observed. This means that, currently, user requests should be “processed correctly” and, therefore, returning to normal behavior in the services.
Amazon working to resolve a global Amazon Web Services outage affecting Fortnite, ChatGPT and Perplexity
— Portaltic.es (@Portaltic) October 20, 2025
“Even if requests begin to be processed correctly, there may be additional latency and some services will have a backlog of work, which could take longer to fully process,” Amazon warned, while stressing that they were still working on the backlog of requests in the queue regarding this incident.