is youtube down — Recommendations glitch behind global outage revealed
Millions of viewers worldwide saw the video platform behave strangely on Tuesday evening as feeds, homescreens and some login flows failed to load. The company traced the disruption to its recommendations system and pushed fixes that restored service across its apps and connected services.
What triggered the outage and how it affected users
The interruption began around 8: 00 p. m. ET when users started encountering empty homepages, missing suggested videos and problems accessing content across mobile and desktop apps. The company confirmed the root cause was a fault in the recommendations system — the algorithmic layer that surfaces suggested videos based on viewing history and signals.
With recommendations failing, the platform could not populate home feeds, and related surfaces such as the music and kids-focused apps were similarly affected. The disruption also touched a subset of connected television logins, with some subscribers unable to authenticate while the broader issue persisted.
Crowdsourced tracking services recorded a dramatic surge in error reports, totaling more than 1. 6 million entries over a 24-hour window tied to this event. Roughly half of those reports originated in the United States, though the outage produced visible errors for users across multiple continents. In some regions, attempts to load the site produced only a blank main area with a sidebar and search box visible.
Timeline: detection, fixes and full recovery (ET)
8: 00 p. m. ET — Initial spike in user complaints and error reports. The platform began notifying users its official communication channels that teams were investigating widespread playback and feed issues.
Shortly after the first notice, engineers narrowed the problem to the recommendations subsystem. That service is responsible for assembling the carousel of suggested clips, which also feeds into app homepages and specialty apps. With that pipeline disrupted, many surfaces could not present content.
Within the hour — a partial restoration returned the homepage for many users, though engineers continued work on a complete fix. Operators also flagged a related issue preventing some people from signing into the TV app; work to resolve those authentication failures ran alongside the recommendations repair.
10: 15 p. m. ET — A final status update announced that the recommendations system had been fixed and that all platform surfaces, including the web, mobile apps, music and kids offerings, and TV apps, were back to normal. Teams monitoring service stability noted a return to expected traffic patterns and a steep decline in new error reports.
10: 26 p. m. ET — Additional post-recovery notes were shared to summarize the incident and close the active incident record.
What this outage means and what users should expect next
The incident highlights how a single automated subsystem can cascade and affect many parts of a complex service. Recommendations play an outsized role in what users see first, which meant the visible symptom was a largely blank experience rather than a complete platform failure.
For most users, normal service has returned and playback and personalized feeds are operating as before. If issues persist, restarting devices, updating apps or waiting for cached elements to refresh can help. The company has indicated it will review the event to reduce the chance of recurrence and improve detection for similar algorithmic faults.
High-profile outages have become more common as large services rely on tightly integrated systems. For anyone still affected, patience and basic troubleshooting are the immediate steps; for the rest of the internet, perhaps it's a reminder to step away from screens and touch some grass.
