If the internet was alive and conscious, Thursday afternoon was its version of a full-blown panic attack. From Spotify randomly forgetting one’s playlist to Gmail ignoring you, core segments of the web just collectively decided to go for a coffee break, all at the same time. Cloudflare blinked, Google Cloud stuttered, and even Amazon Web Services was accused of sneezing (though it claims it was okay). For an instant, it seemed like the online world ran into a pothole.

Incident tracker Downdetector.com noted there was a sudden increase in reports of incidents during early to mid-afternoon, hitting both backend operations and user-facing platforms. The broad technical disruption, which started at approximately 2:45 p.m. EDT, was somewhat resolved by evening, but not before exposing the fragile web of today’s internet infrastructure.

Cloudflare Accuses Google Cloud Outage

Cloudflare, a fundamental element of the internet’s plumbing that’s famous for directing and shielding tens of millions of web requests every second, reported it was seeing service outages starting mid-afternoon. 

In a 2:46 p.m. EDT status update, the company noted

“a number of services suffer intermittent failures.”

More than two hours later, Cloudflare announced all services were back to normal by 4:57 p.m. EDT. Cloudflare blamed the disruption on an outage at Google Cloud, explaining that

“a limited number of services at Cloudflare use Google Cloud and were impacted.”

Global Disruptions Reported by Google Cloud

Google Cloud also experienced its own issues at the same time. The cloud behemoth reported widespread disruptions to numerous of its services globally and started releasing status notifications confirming the outages. Google Cloud at 5:23 p.m. EDT announced most of the impacted products had recovered. Google confirmed that they were investigating disruptions to some services, but would not provide concrete technical information.

The cascading effect of the Google Cloud outage was clearly evident, as many other platforms based on or dependent on its infrastructure started to break down.

Amazon Web Services Dismisses Outage despite Reports of Spiking

Downdetector also saw a spike in Amazon Web Services (AWS) outage reports, reaching as high as almost 6,000 by 2:49 p.m. EDT. AWS’s own service status page reflected no issues, though Amazon pushed back on the reports as being inaccurate, stating that AWS services are “operating normally.” Even so, the spike in reports on the part of users indicated at least an apparent degradation of service on applications powered by AWS.

Other Affected Platforms

The wave of outages issues also swept across several high-profile online services, such as Spotify, YouTube, Google Meet, Gmail, Discord, Snapchat, and Etsy. According to Downdetector, all suffered major outages within the same period, although most of the services were back to normal by 3:55 p.m. EDT.

Interestingly, social media giants like X (formerly Twitter), Facebook, and Instagram seemed to avoid the full force of the outage. Outage reports for these sites were minimal, with all of them collectively measuring only a few hundred incidents.

Elon Musk’s X Glitches

Though Thursday’s outage mostly spared social media sites, X (formerly Twitter), has also seen its own reliability issues recently. In May, there were tens of thousands of reports of outages for X during several incidents. Musk himself referred to the site’s issues, stating that he was

“back to spending 24/7 at work and sleeping in conference/server/factory rooms.”

He added that X needed

“major operational improvements.”

Infrastructure Dependence

The incident emphasizes the crucial function of cloud infrastructure providers like Google Cloud, Amazon Web Services, and Microsoft Azure. Based on the 2024 figures from market research company HG Insights, Google Cloud now holds approximately 28% of the U.S market share among leading cloud service providers. This was not the first outage of this year that Google Cloud experienced, a total of four product-specific outages were reported in 2024. Thursday’s broad interruption did touch many more regions and services, highlighting just how reliant the internet ecosystem is on a handful of central players.

Thursday’s shutdown, however temporary, was a shining sign that reminded us just how heavily our everyday infrastructure, commerce, communication, entertainment, and even crisis management, relies on a small number of cloud behemoths. When one falters, the ripple can be felt throughout the world. 

As tech firms pursue scale and governments grapple over regulation, the user in the middle is left to wonder and ask why their playlist will not load or where their client meeting went. In the future, the industry needs to invest not only in redundancy but in transparency. We need a digital landscape that’s not just vast and efficient but resilient and accountable when it fails.