Talks are ongoing between President Donald Trump and Chinese leader Xi Jinping on a potential TikTok deal. The Wall Street Journal is reporting that TikTok engineers will re-create a set of content-recommendation algorithms for the app in the US, using technology licensed from TikTok’s parent ByteDance. Such an arrangement could assuage concerns about national security. On April 24, 2024, the U.S. enacted the Protecting Americans from Foreign Adversary Controlled Applications Act, which requires TikTok to be sold by its parent company, ByteDance, to non-Chinese owners or face a nationwide ban. The Supreme Court upheld the law in January 2025.
The TikTok Algorithm and Engagement
Users on digital platforms face an overabundance of viewing choices. Unlike in the past where a consumer had to drive to a physical store and spend time browsing the limited inventory, now we have seemingly endless choices. Recommendation algorithms on various platforms, be it Netflix, TikTok, or YouTube — control how we navigate the choices made available by on-demand streaming media, since there is too much content for any one user to filter through. The digital overabundance of choices increases our dependence on algorithmic filters to curate and sort our entertainment preferences. This is where TikTok has an edge over its competitors. TikTok’s recommendation algorithm is hyper-personalized. The platform tracks incredibly detailed engagement patterns, such as video interactions (like, share, how long you watch a video), scrolling behavior (speed and pattern of navigation), your search history, the time spent on the app and your content creation patterns. TikTok also gathers very detailed location information, ranging from browsing habits and GPS location to interactions within apps, including precise GPS coordinates (if enabled), location inferred from IP address and time zone information. All this information is analyzed to personalize a “For You” page, predict what content consumers engage with, target advertisements and to identify trending content and creators. The sheer volume and granularity of this data collection allows TikTok to build incredibly detailed user profiles, enabling their highly addictive recommendation algorithm while raising significant privacy concerns about how this information is stored, processed, and potentially shared. A new ownership structure could allow American supervision of the algorithm.
TikTok’s Role in Ecommerce
Traditional e-commerce platforms are goal oriented and assume customers know what they want and come to shop with a specific purpose. E-commerce platforms such as Amazon implicitly assume that users have a clear purchase intent, and success is measured by quickly users can find and buy what they need. TikTok on the other hand enables product discovery while users are watching entertaining videos, often impulse-buying items they weren’t searching for. The TikTok model treats shopping as entertainment and focuses on helping users discover things they didn’t know they wanted. The platform suggests items based on trends, lifestyle, or visual appeal, where shopping becomes a leisure activity, not just a transaction. Products are discovered through influencer content, user reviews, and social features. This shift reflects changing consumer behavior, especially among younger demographics who prefer discovery and entertainment over traditional product research. It also allows platforms to introduce users to a much broader range of products than what consumers might have searched for on their own. Indeed, it is estimated that TikTok users in the U.S. have spent close to $41.3 million in the app to date, or about 55 percent of the global total, placing the US first in terms of user revenue for the app.
Why Does Ownership of TikTok Matter?
Most users of digital platforms are completely in the dark about the extent to which platforms monetize user interactions, and in turn, how nudges and recommendations from platforms impact aggregate behavior. Social media companies primarily monetize through advertising, which relies heavily on user data to target ads effectively. Users’ personal information, behavioral patterns, and social connections are essentially the product being sold to advertisers. When users lack control over their data, they can’t make informed decisions about their digital footprint or protect sensitive information from being used in ways they didn’t anticipate. Such concentrated data ownership gives platforms significant influence over information flow, user behavior, and even democratic processes, as seen in lawmakers concerns about election interference and misinformation.
Most major platforms now offer some data control options, largely due to regulations like Europe’s GDPR and California’s Consumer Privacy Act (CCPA), that allow for data portability, deletion rights, and enhanced privacy settings. Under such regulations, companies are also required to publish transparency reports disclosing what data they collect and how it’s used, though these are often lengthy and technical. Despite these controls, true data ownership remains elusive because terms of service often grant platforms broad rights to use consumer data. Data deletion doesn’t necessarily remove copies used for analytics or shared with third parties. The complexity of modern data processing makes it difficult to track all uses. The key requirement from US legislation is that ByteDance must divest its control, but the specific ownership structure of the replacement entity remains flexible. It is worth noting that ByteDance’s ownership is already diversified, with 60% owned by global institutional investors such as BlackRock, General Atlantic, and Susquehanna International Group.
How AI Could Threaten TikTok’s Dominance
As AI-powered search becomes more sophisticated, it could pull users away from TikTok’s discovery-driven model. While the younger generation prefers TikTok to typing queries on Google, AI assistants could reclaim this behavior by providing more personalized, conversational answers. AI also threatens TikTok’s creator economy by potentially flooding the platform with synthetic content.
TikTok’s core strength is its recommendation algorithm, but AI advancements could level this playing field through better personalization and cross-platform AI. New AI models can analyze user behavior across multiple platforms for superior targeting. AI could also enable entirely new types of social platforms and social interaction, such as the ability to personalize content to end users rather than just recommending existing content. Another way that AI could shape digital platforms is through enabling synthetic social experiences wherein virtual influencers and AI companions that provide entertainment without human creators. Finally, an era of integrated AI assistants could emerge where social media, search, and AI assistance converge into one experience.
While the final details of the TikTok deal are being ironed out, it is worth pointing out that AI could erode TikTok’s key strengths in content creation and personalization. Here, again, though, TikTok could also have an edge in that its AI models are equipped with detailed behavioral data gathered from its users. Data could yet be the new currency in the new AI-enabled landscape.