US TikTok Disruption Linked to Data‑Center Power Failure Amid Fresh U.S. Ownership

An operational incident beginning early Sunday left many U.S. TikTok users unable to upload videos, see fresh content reliably, or use interactive features; the platform’s newly formed U.S. data‑security unit attributed the problems to a power failure at a domestic data center. Users reported broken For You feeds, failed uploads, slow or missing comments, and inconsistent direct‑message behavior that sometimes blocked single words while allowing identical words in full sentences, which fed confusion and conspiracy claims. Independent monitoring services recorded a sharp spike in outage reports consistent with a regional infrastructure failure rather than a global moderation change. The timing — days after a $14 billion U.S. spin‑out that placed Oracle, Silver Lake and Abu Dhabi’s MGX in operational roles while ByteDance retained a minority stake — intensified scrutiny and made the outage politically sensitive. Technically, the symptoms point to partial degradation across recommendation, publishing and messaging subsystems rather than a clean offline event, complicating troubleshooting and public explanations. Public reaction was swift: a measurable number of users sought alternatives and at least one emergent competitor, UpScrolled, recorded tens of thousands of installs over the weekend, briefly overwhelming its onboarding capacity and exposing gaps in its infrastructure. That surge illustrates a common pattern: governance or reliability shocks at a dominant platform can produce immediate but often transient migration, testing rivals’ ability to scale and retain new users. For the new U.S. owners, the immediate exposure is reputational — infrastructure faults can be misread as censorship in a highly charged regulatory environment — and operational, as partners and advertisers demand stability guarantees. Resolving the technical fault is necessary but insufficient; the company needs clearer incident timelines, richer telemetry for public verification, and demonstrable redundancy so a single data‑center failure cannot cascade into user‑facing moderation narratives. Regulators and lawmakers watching the transition may press for contractual resiliency clauses and independent audits, increasing compliance and reporting burdens. Advertisers and creators may see short‑term engagement dips and will look for assurances or compensation if outages become frequent. In sum, the incident highlights both the immediate operational fixes required and the broader governance communications needed to support a politically charged ownership transition.