The Day the US Controls TikTok's Brain: Will China Let Go of the Algorithm? - Oracle Protects, and the US Decides?

The White House has declared a new framework to oversee TikTok's operations in the U.S., stating that "the U.S. will control the algorithm, with six out of seven board seats held by Americans." Oracle will handle data and security, aiming to sustain the app while avoiding legal risks. However, China has not officially agreed. Is the agreement "imminent," or is it a smokescreen for political negotiations? Analyzing the implications for creators, advertisers, and U.S.-China relations. Barron's

1. What Has Been "Agreed" — Clarifying the Public Line

White House spokesperson Caroline Levitt explained that (1) the U.S. will have control over TikTok's recommendation algorithm, (2) Americans will occupy six of the seven seats on the new oversight board, and (3) Oracle will handle the protection and security of U.S. data. The government's explanation aims to prevent "invisible interference" in the domestic discourse space. Barron's

On the other hand, the Chinese government and parent company ByteDance have not officially acknowledged the agreement on key points, including the handling of the algorithm. Even in recent U.S.-China leader calls, no assurances have been publicly given by the Chinese side. Barron's

2. Political and Legal Context: Why Now?

This framework is an extension of the bipartisan law enacted in 2024, which demanded "sale or shutdown" (TikTok Regulation Act). Reports indicate that even after the Supreme Court supported the law, President Trump repeatedly postponed the enforcement deadline, most recently extending it to the end of the year (December 16) citing the "broad outline of an agreement." Politico

Diplomatically, the U.S. and China have approached a "principle agreement" on economic issues, including TikTok, with mentions of rescheduling a Trump-Xi meeting. However, the content and process of the "principle agreement" remain unclear, and cautious views persist. Reuters

3. Technology and Governance: The Reality of Algorithm "Sovereignty"

3-1. What Does Algorithm Management Mean?

"The U.S. controlling the algorithm" can be interpreted as the U.S. holding decision-making power over the update, evaluation, and distribution control of the recommendation model. The White House emphasizes containing "influence operations," but academically and practically, it is divided into the following layers.

• Data Layer: Keeping U.S. user data within the U.S. (Oracle's secure environment).

• Model Layer: Access/change rights to the code and parameters of the recommendation model (representation learning, ranking).

• Operation Layer: Designing A/B tests, applying safety policies, emergency shutdown authority, etc.

According to the government explanation, Oracle will continue to handle the infrastructure, and the oversight board will be predominantly American, but the delineation of the model's intellectual property (ownership of source code, export control) and transfer of training data remains at the heart of negotiations. Barron's

3-2. The "U.S. Version of TikTok" Theory and Implementation Risks

Some reports mention the possibility of the app being "split" between the U.S. and international versions. If learning and operation occur only within the U.S., concerns arise about model performance degradation (reduced data volume and diversity), delays in feature development (discrepancy with global development lines), and revenue impact from ad optimization (signal loss). Barron's

4. Economic and Industrial Impact

4-1. Creators/Advertisers

If U.S. algorithm management increases safety and transparency, major advertisers focusing on brand safety may return and expand their placements. On the other hand, the decline in recommendation accuracy due to split operation could negatively affect viewer retention rates and CPM. Creators will likely further pursue multi-homing with YouTube Shorts and Reels to prepare for revenue volatility (sudden changes in distribution behavior).

4-2. Big Tech Competition

For rivals, TikTok's "domestication" presents a temporary opportunity for offensive. Meta, in particular, can aim to capture DAU/stay time with Reels, and Google with Shorts. Conversely, if TikTok "clears" U.S. regulations, it could gain advertisers' trust as a regulation-resistant platform in the long term.

5. Diplomatic and Security Perspectives

5-1. U.S. Intentions

The U.S. has been exploring a "third way" to "suppress influence operation risks without stopping the platform." Algorithm management and a U.S.-led board are the core of its political accountability. Politico

5-2. China's Calculations

The current situation, where China has not acknowledged the agreement, can be read as a stage where it is measuring alignment with domestic norms such as "source code export regulations" and "national security review." It is natural for China to be cautious about measures that could be perceived as a "complete transfer of sovereignty." Barron's

6. Points Still in the Fog (Risks)

1. Final Contract Signing and Timeline: The U.S. side claims it is "almost in the final stage," but there is no official confirmation from China. Barron's

2. Algorithm Intellectual Property Boundaries: What are the extents of viewing, operational, and modification rights?

3. Emergency Intervention Authority: Governance of feed control during elections and emergencies.

4. "Alternative Scenarios" for U.S. Shutdown: Shutdown measures and creator protection strategies if the agreement falls through.

5. Linkage with U.S.-China Diplomacy: Potential "links" with other issues such as trade, semiconductors, and Russia policy. Reuters
   
   

7. Reactions on Social Media: The "Divide" of Pros and Cons

7-1. Skepticism and Cynicism

On Reddit and Hacker News, skepticism is prominent, with questions like "Can the U.S. really control the algorithm?" and "Will we end up having to create a 'U.S. version of TikTok' from scratch?" In one thread, an article headline stating "China retains the algorithm" spread, with voices questioning the effectiveness of the agreement rising to the top. Reddit

7-2. Voices Expecting Safety

On the other hand, many posts view localization and strengthened oversight positively. Comments such as "If this protects young people's data, it's welcome" lined up on local stations and news pages' Facebook posts, with supportive reactions following. Facebook

7-3. Creators' Concerns

Among creators, concerns like "If recommendations slow down, it will hit revenue directly" and "Will features be delayed only in the U.S.?" are shared, with posts visible about strengthening parallel posting to YouTube/Instagram as a backup (from the discussion and quoted links in the above threads). Reddit

8. Practical Advice for the Business Side

• Q4 Campaigns: Flexibilize bid management and strengthen portfolio diversification to Reels/Shorts.

• Brand Safety: Use category exclusions and third-party verification together.

• Creators: Prepare for funnel diversification (email lists/Discord/other short-form) and hedge revenue.

• Legal and Policy: Routine tracking of primary information (White House announcements, Department of Commerce announcements, ByteDance statements, Oracle disclosures).
  
  
  

9. Conclusion: The Agreement is "Almost Landed," but the Real Issues are Yet to Come

The declaration of "U.S. algorithm management" appears politically as a "victory declaration." However, whether it can be translated into the technical, operational, and intellectual property layers is another matter. Can the U.S. version's local operation maintain a global standard experience? To what extent will the Chinese side consider it a transfer of sovereignty? The details of implementation and protocols of mutual trust are the real battlegrounds. Barron's