Key Points

Previous concerns about massive user departures from TikTok's US joint venture now appear overstated. Application uninstallations saw a brief surge, but usage time and download numbers quickly recovered. Data indicates most American users have not noticed substantial operational changes at TikTok.

January 27, 2026 - A sign in front of the TikTok building in Culver City, California. Recent data shows that TikTok's US joint venture has navigated through a turbulent transition period with its user base largely intact. Earlier predictions of a "mass user exodus," fueled by service disruptions and censorship concerns, now seem significantly exaggerated.

According to an investigation by market intelligence firm Sensor Tower, although application uninstallations spiked temporarily after TikTok announced the formation of its US joint venture on January 23, the number of daily active users in the United States remained at approximately 95% of the level seen during the week of January 19-25.

The entity, officially named TikTok USDS joint venture, was established to comply with an executive order from US President Trump mandating the separation of TikTok's US operations from its parent company, ByteDance.

Under the agreement, ByteDance retains a 19.9% stake in TikTok's US business. Oracle, Silver Lake Capital, and Abu Dhabi-based investment firm MGX each hold 15% stakes, with the remaining shares distributed among several other companies.

Following the joint venture announcement, users quickly expressed dissatisfaction with the new ownership structure.

The deal prompted widespread skepticism, with prominent figures, including Vermont Independent Senator Bernie Sanders, raising concerns about "cronyism" related to the involvement of Oracle co-founder and Chief Technology Officer Larry Ellison.

After the joint venture announced that Ellison's Oracle would "retrain, test, and update the content recommendation algorithm based on US user data," online speculation surged. Theories suggested TikTok might begin mining user data or promoting content supportive of Trump's policy positions.

These concerns peaked on January 25, with users alleging that TikTok was suppressing content critical of controversial actions by US Immigration and Customs Enforcement (ICE) and blocking trending terms like "Epstein" on the platform.

Last month, CNBC verified that messages containing the word "Epstein" triggered error notifications, but could not independently confirm broader allegations of political censorship.

A spokesperson for the TikTok joint venture stated in a January interview with CNBC that the platform had not banned the use of the name "Epstein" in messages and was investigating issues reported by some users.

CNBC reached out to the White House and TikTok for comment but had not received responses by the time of publication.

User Engagement Metrics Unchanged

Although TikTok attributed last month's glitches to a power outage, Jim Johnston, a partner at law firm Davis+Gilbert, believes these disruptions "undoubtedly affected how and what content was delivered, even if there was no specific intent or motive."

However, despite some users pledging to boycott the platform over alleged political suppression, US user engagement data shows little evidence of a large-scale departure.

Sensor Tower data indicates that the average daily usage time for US users dipped to 77 minutes during the week glitches were reported but has since recovered to approximately 80 minutes.

Furthermore, while uninstallations surged following the glitch reports, they subsided in the subsequent week, indicating a short-term fluctuation rather than a sustained boycott.

"The short-term spike in uninstalls was likely driven by users attempting to troubleshoot app issues," Abraham Yusuf, Senior Insights Analyst at Sensor Tower, told CNBC. On January 25, the number of users who uninstalled and then reinstalled the app the same day surged by over 70% compared to the previous day.

Yusuf acknowledged that data shows "overall usage was slightly impacted" in the weeks following the joint venture announcement, but there is no clear sign of a structural shift in user trends. Many platforms viewed as alternatives to TikTok have also struggled to maintain user interest.

According to Sensor Tower, the social platform UpScrolled, which promotes having no algorithmic filtering (i.e., no "shadow banning"), saw new downloads surge by approximately 770% week-over-week for the period January 26-February 1, with over 955,000 new downloads in the US.

However, in the following week, UpScrolled's new downloads plummeted by about 80%, adding only roughly 191,000 new users. In comparison, TikTok garnered 870,000 downloads during the week of January 26-February 1 and approximately 800,000 the following week.

Similarly, other alternative platforms like Skylight Social and Red Note saw their new downloads decline by 96% and 33% respectively in the week following January 26 compared to the previous week.

Lack of Evidence for Mass Exodus

Sensor Tower's user data fundamentally indicates that, beyond anecdotal reports, most US users have not perceived substantial changes in TikTok's operations, at least not enough to significantly alter user behavior.

"Claims of a massive TikTok user exodus now appear overblown," Kelsey Chickering, Principal Analyst at Forrester Research, told CNBC. "Feedback from most users suggests the app experience is largely unchanged – the algorithm hasn't undergone major adjustments, and the user experience remains solid."

Although some US users might perceive changes in how TikTok's algorithm functions, Johnston believes that "any partial modification in content recommendations is simply because the dataset has changed" – referring to the joint venture's announced plan to retrain the algorithm using US data.

Nevertheless, while analysts have found no evidence yet that TikTok's new US owners are manipulating the platform for their own benefit, this is not a definitive conclusion.

Johnston pointed out that TikTok's new user agreement includes at least three notable changes: the platform can collect precise location data from enabled devices, gather data on user interactions with in-app AI tools, and explicitly integrates with advertising networks.

Johnston stated that, while there is no concrete evidence, it remains technically feasible to adjust TikTok's algorithm to amplify or diminish the reach of specific content in recommendations.

Chickering added that under the new ownership, TikTok has greater control over the content appearing in US users' feeds, and this control represents both an opportunity and a risk for the platform.

"If content moderation begins to show political bias, or if misinformation is not handled appropriately, the platform could face backlash from both users and advertisers," Chickering said. "We've seen this before: the rebranding of Twitter to X serves as a recent case study reminding us how quickly user trust can erode."

But for now, the initial discontent that troubled TikTok's US user base during the early days of the new ownership has largely subsided.

As Chickering noted, "We've repeatedly seen that if the product works well, users often don't care who owns it."