Character AI went down Monday afternoon, leaving users unable to reach the service after 2 p.m. EST and generating thousands of problem reports.
Monitoring site DownDetector showed more than 2,000 reports of issues on Monday, and its tally rose higher during the outage window — more than 3,000 users had reported problems with the platform as of 11:44 a.m. PT. The company’s mobile app accounted for the bulk of complaints: 63 percent of the reports flagged app failures.
Users also reported trouble on the website, and a smaller number were unable to log in at all, spreading the disruption across both mobile and browser access.
DownDetector identified the incident as a service outage Monday afternoon and registered the surge in user reports that followed. That measurement comes from aggregated status reports collected from multiple sources and reflects the volume of users who posted problems during the window.
Character.AI’s public status checker offered no timetable for restoration. The status message said the company was currently investigating the issue, leaving users with no clear estimate for when normal access might resume.
The scale of the reports — multiple thousands within hours — marked the outage as more than an isolated hiccup. With 63 percent of reports tied to the app, mobile users bore the brunt of the disruption, while web and login failures expanded the impact to people trying to reach the service from different devices.
For users, the immediate consequence was interruption: sessions cut off, characters and conversations inaccessible, and some who tried to sign in blocked from entering the platform. For a system that depends on continuous availability to keep conversations and user workflows intact, those interruptions create a clear usability break for everyone trying to use it in real time.
The friction in the timeline stands out: the outage began after 2 p.m. EST, yet the public status entry remained at the investigative stage while reports continued to climb. That mismatch — a confirmed outage alongside a single-line status update that gives no repair target — leaves the scale of the problem documented but the path to recovery unclear.
Company engineers, judging by the posted status message, had acknowledged the event and were looking into causes, but the public-facing update did not identify a root cause or an estimated fix time. That gap is consequential because the outage affected multiple access methods rather than a single client, suggesting either a backend failure or a broader infrastructure hiccup.
Readers should expect the central unanswered question to be when Character AI will restore full service. The company’s statement that it is investigating is the latest official step; beyond that, there is no published timetable for a fix and no technical detail released about the failure mode.
Until engineers provide an updated status with a resolution estimate or rollback details, users will need to rely on the service’s status page and monitoring sites for real-time indicators of recovery. The outage—measured in thousands of reports and concentrating on the app—will remain unresolved publicly until the company moves from investigation to remediation and posts a clear update.
