TikTok closed a deal with the US on Jan. 22 to allow the app to stay operational in the country. This deal has followed a multi-year scuffle between the US and China that began during President Trump’s first term in the White House when he initially floated the idea of a ban on TikTok in the US.
The ban gained traction in 2024, when President Biden signed a law stating that ByteDance, TikTok’s owner, must either sell TikTok or face a ban on US operations of the app due to concerns that China would use the app to steal US users’ data. The app went offline for about 12 to 14 hours before being restored when then president-elect Trump pledged to postpone the ban.
This past September, President Trump announced that he had reached a deal with China to keep TikTok operating in the US, and official agreements were signed by American and global investors in December. Under these agreements, a new business called TikTok USDS Joint Venture LLC will control US user data, apps, and the algorithm to protect national security.
The Joint Venture is owned by a number of US-based companies and investors who hold an 80.1% stake, while ByteDance has retained a 19.9% minority stake. The three key investors are technology company Oracle—headed by Trump ally Larry Ellison—private equity firm Silver Lake, and MGX, an Emirati investor in AI and technology.
TikTok’s content recommendation algorithm is what has caused the app’s increase in popularity over the years since its launch. Other platforms, such as YouTube and Instagram, have tried to replicate it with similarly-formatted programs—YouTube Shorts and Instagram Reels, respectively—but users have reported that the content curated by their algorithms is not personalized to the extent of the TikTok algorithm.
ByteDance has developed the TikTok algorithm with something called “interest signals” instead of “social graph” (which is what Instagram’s Meta uses) to train their algorithm. “Interest signals” are essentially digital breadcrumbs that allow the algorithm to track users’ changing preferences and interests over time as to personalize their content and advertisements; this includes what content users may like to see at different times of the day.
TikTok also regularly recommends videos and content that doesn’t necessarily align with user interests, which is another way that they infer how and when their interests in different content is shifting.
According to TikTok, “The Joint Venture will retrain, test, and update the content recommendation algorithm on U.S. user data. The content recommendation algorithm will be secured in Oracle’s U.S. cloud environment.”
This has raised concerns that the shift to Oracle controlling the algorithm instead of ByteDance will change and depersonalize it, because Oracle will be retraining the algorithm as a way to maintain national security and implement “content moderation rules,” putting Oracle in charge of what content users in the US will be able to see on the app. This has led to unanswered questions about how content might be censored, and how it will affect US users interacting with global users’ content and vice versa.
On Jan. 22, US TikTok users were presented with a pop-up screen informing them of TikTok’s new Terms of Service and Privacy Policy. One bullet point in the pop-up stated that they will collect “New types of location information (including device geolocation)…”
This shift in wording is one of the more stark changes to the app, because the previous privacy policy has stated that the app did not collect precise GPS information. The fact that the use of geolocation is explicitly stated has also raised concerns over personal security.
Under the “Information You Provide” section of TikTok’s updated Privacy Policy, it is stated that TikTok will collect users’ “racial or ethnic origin, national origin, religious beliefs, mental or physical health diagnosis, sexual life or sexual orientation, status as transgender or nonbinary, citizenship or immigration status, or financial information” if it is disclosed under survey responses or within ones’ user content.
These two aspects of the new update have raised concerns specifically in relation to recent news of ICE using surveillance apps such as Tangles and Webloc, which are used to track the cellphone activity of entire neighborhoods.
Tangles has been used in the past to scrape sites such as Twitter—now known as X—and make that information available to authorities. Though, using Tangles along with Webloc allows ICE to use location data to search a map-style interface for phones and people they’re interested in. Webloc allows users to circle a location on the map, and it provides all the location data and phones it has for that location, further allowing users to track those phones to new locations.
It is not known exactly what all of this means for TikTok users in the US. What we do know is that the algorithm will face changes (exactly what those are is unknown); there will be stricter rules around AI-generated content, including that users are responsible for labeling it as such; there will be a restricted, limited experience for children under 13; the app will collect more data and it will be using increased location tracking (though a TikTok official has stated that users can opt out of that); and there might be shifts in political content moderation.
