- U.S. investors, including Oracle and Silver Lake, are positioned to acquire 80% of TikTok’s U.S. operations.
- Oracle will oversee TikTok’s prized recommendation algorithm, licensed from ByteDance but retrained on U.S. user data.
- ByteDance will retain a minority stake (≤20%) while leasing algorithm technology to the U.S. entity.
- The deal must pass regulatory approval in both Washington and Beijing, leaving its future uncertain.
- Experts warn U.S. government oversight could introduce its own risks for content neutrality and free expression.
Proposed deal puts algorithm retraining on American data at the heart of national security assurances.
After months of tense negotiations, a proposed deal has emerged that would hand majority ownership of TikTok’s U.S. operations to American investors. Under the framework, Oracle, Silver Lake Partners, and media figures such as Rupert and Lachlan Murdoch are expected to take a collective 80% stake in the new venture. ByteDance, TikTok’s Chinese parent company, would retain no more than 20%.
The move aims to meet the divest-or-ban mandate signed into law last year while also preserving the platform for its 170 million American users. A White House official described the arrangement as a “compromise that meets both the requirements of U.S. law as well as the domestic law in China.”
The Algorithm at the Center of the Deal
TikTok’s recommendation algorithm—the technology that dictates what appears in users’ feeds—has become the most contentious issue in the negotiations. Oracle is set to serve as the “security provider,” charged with overseeing and securing this system.
The algorithm will be licensed from ByteDance but retrained exclusively on U.S. data. All American user data will be stored on U.S.-based Oracle servers, expanding the Project Texas safeguards already in place.
“The details are really going to matter to understand whether this addresses the national security concerns,”
noted Samm Sacks, senior fellow at Yale Law School’s Paul Tsai China Center.
While this arrangement may assuage some fears of Chinese government influence, it does not eliminate questions about who ultimately controls access to data and how recommendations may be influenced.
Regulatory Hurdles in Two Capitals
Approval is still required from both Washington and Beijing. President Donald Trump has signaled support, promising to sign an executive order declaring the deal compliant with national security needs. The order is expected to include a 120-day window to finalize implementation.
Beijing, however, has not yet officially confirmed its stance, despite signals that it may allow the licensing of the algorithm. Without approval from China’s regulators, the deal could stall.
National Security Versus Free Expression
The Electronic Frontier Foundation (EFF) has raised concerns that U.S. government involvement in algorithm oversight could create its own risks.
“If the concern had been that TikTok could be a conduit for Chinese government propaganda … people can now be concerned that TikTok could be a conduit for U.S. government propaganda,”
stated David Greene and Eva Galperin of the EFF.
This tension highlights the broader debate: securing the app against foreign influence may inadvertently open the door to domestic political influence over one of the most powerful content engines in the world.
Implications for Users and Creators
For everyday users, the deal promises continuity—TikTok will not go dark in the United States. But the retraining of the algorithm on U.S. data could gradually reshape the content experience. Creators and advertisers may see subtle shifts in recommendation patterns, which could affect reach, engagement, and monetization.
Analysts warn that any perceived changes to the feed could trigger significant shifts in user behavior.
“Material (or even perceived) changes to the content, algorithm or app policies could prompt massive shifts in user behaviour,”
said Jasmine Enberg, principal social media analyst at eMarketer.
A Test Case for Global Tech Governance
The TikTok saga illustrates the collision of technology, national security, and geopolitics. It also sets a precedent for how governments may handle future disputes over digital platforms with global reach.
The final terms—particularly around algorithm licensing, oversight, and data flow—will shape not only TikTok’s U.S. future but also the broader conversation around sovereignty and technology governance.
As the deal moves toward implementation, the unresolved question remains: can retraining an algorithm and shifting ownership truly neutralize the geopolitical concerns surrounding TikTok? Or does the compromise simply replace one set of risks with another?
