Users of X (formerly Twitter) reported fresh disruptions on March 23, 2026, with the problems surfacing in the late afternoon and lasting at least into the early evening hours in the U.S.

Multiple posts and complaints pointed to trouble accessing the platform, particularly around 4:30 p.m. EST. For many people, the site didn’t load reliably, and log-in attempts often failed or stalled.

Downdetector, a real-time service that tracks online outages based on user reports, showed a clear spike in complaints during the same window. The pattern suggested a widespread issue rather than isolated problems on individual devices or networks.

H2

Why X Outages Hit Users Hard Around Login Attempts

The data indicated that the biggest wave of frustration was tied to the X website experience. According to Downdetector statistics, roughly 60% of affected users said they couldn’t successfully get onto the site.

That category typically includes a range of issues—pages failing to load, timeouts, or sessions that won’t establish correctly. In practical terms, it means users trying to scroll, refresh, or open their feeds may have been met with error screens or endless loading indicators.

Another 17% of users reported problems with the mobile app. For those users, the issue may have presented differently—such as the app crashing, failing to connect, or becoming unresponsive while attempting to fetch content.

Meanwhile, 14% of reports were specifically tied to users being unable to get past the login screen. Login failures are often among the most disruptive outages because they prevent people from even reaching the platform, regardless of whether the feed or messaging components are otherwise functioning.

Downdetector’s breakdown also helped highlight that the outage wasn’t limited to a single symptom. Instead, users experienced a mix of access failures across site login, app connectivity, and general website performance.

At the time of reporting, there was no official timeline from X on when the issues would be fully fixed. That uncertainty can be especially frustrating for people who rely on the platform for time-sensitive communication, news updates, or work-related announcements.

For users trying to troubleshoot, typical strategies—like switching networks, clearing cached data, or restarting the app—may offer temporary relief, but they do not address server-side problems. When outages are broad, individual fixes often can’t fully restore access until the platform’s systems stabilize.

H2

No Clear ETA as Downdetector Tracks Spike in Reports

Downdetector’s monitoring suggested the spike began shortly before 4:30 p.m. EST, then continued as more users attempted to access X. Those additional attempts likely increased the number of reports, especially among people who repeatedly retried login or refresh operations.

It’s also worth noting that outage severity can fluctuate quickly. Even when a platform isn’t completely down, partial disruptions—such as slow authentication, unstable page rendering, or intermittent connection drops—can generate a high volume of user complaints.

The reporting distribution—majority website access issues, followed by app problems and login blockages—aligns with what many users typically see during authentication or routing problems. In these cases, a platform might allow some users through while leaving others stuck at the same step.

For creators, businesses, and journalists who depend on real-time engagement, even short disruptions can have noticeable consequences. Posts may not load, replies can fail to send, and scheduled content may be delayed if publishing tools don’t connect properly.

As of this update, there was still no announced estimate for resolution. Users were left watching for signs of improvement—such as successful logins, pages loading normally, and fewer error messages.

If the outage is being addressed behind the scenes, recovery often happens in stages. First, login systems stabilize, then general page access improves, and finally apps reconnect consistently. Until then, reports are likely to remain elevated.

Still, the fact that multiple issue categories were documented suggests the problem may have been broad enough to affect more than one component of the service. That can slow down recovery, especially if systems require monitoring and verification before declaring stability.

Users are advised to keep an eye on live outage tracking platforms and official updates from X, since those sources usually provide the most direct signals of when service begins to normalize.

Some people may already be able to access the platform, while others continue to face the same barriers. Until the platform confirms repairs and Downdetector reports begin to level off, the disruptions appear likely to remain in place for at least a short stretch of time.