“I think anytime you have [China] with leverage over the algorithm, I think that’s a problem.”
His skepticism arrives as the White House celebrates a deal transferring TikTok’s U.S. operations, valued at $14 billion, to a consortium including Oracle, Silver Lake, and Abu Dhabi’s MGX investment fund. The transaction, approved by President Trump’s executive order on September 25, 2025, gives the new American-controlled entity 80% ownership while ByteDance retains 19.9% and appoints one of seven board members. Yet beneath this restructuring lies a troubling reality: the algorithm that determines what 170 million Americans see daily won’t be severed from ByteDance but licensed from it.
The Technical Fiction of “Retraining”
The proposed deal’s structure reveals its central contradiction. Under the framework, the new U.S. entity would lease a copy of TikTok’s algorithm from ByteDance, which Oracle would then “retrain from the ground up” while monitoring its security. White House officials insist this arrangement removes operational control from foreign adversaries, but Moolenaar’s response cuts through the corporate euphemisms: “I just believe you have to have a new algorithm, and I don’t know that you can reprogram.”
This skepticism reflects broader concerns among cybersecurity experts that algorithms of this complexity cannot simply be audited and declared secure. TikTok’s recommendation engine has been refined through billions of user interactions over years, processing data from 82.2 million daily active U.S. users who spend an average of 53.8 minutes per day on the platform. Creating a genuinely independent system would require rebuilding the entire machine learning infrastructure from scratch, potentially destroying the very feature that makes TikTok valuable.
The licensing arrangement necessitates ongoing technical cooperation, creating multiple potential vectors for influence or data access. Even with Oracle’s oversight, the fundamental architecture including the mathematical principles, optimization strategies, and core machine learning models would remain rooted in ByteDance’s intellectual property. Any meaningful algorithm update or adaptation would flow through this licensing relationship, making complete independence a corporate fiction rather than technical reality.
Analysts initially estimated TikTok’s U.S. operations could be worth between $30 billion to $40 billion, yet the deal values it at just $14 billion according to Vice President JD Vance. This steep discount suggests ByteDance understands the ongoing algorithm licensing creates residual value and control that goes well beyond the 19.9% equity stake Congress’s 2024 law was designed to eliminate.
Trading Foreign Control for Billionaire Influence
The consortium’s composition introduces concerns extending beyond traditional national security considerations. Oracle, which manages data security and algorithm oversight, is not a neutral guardian but a profit-driven technology company led by Executive Chairman Larry Ellison, a prominent Trump supporter and Republican megadonor. Fox Corporation, controlled by the Murdoch family, is also reportedly involved in the deal, raising questions about how the new ownership structure might influence content moderation and algorithmic curation.
This arrangement represents what analysts characterize as the privatization of the attention economy. A platform shaping the information diet of 170 million Americans, disproportionately young users, with 59% of Americans aged 18-29 using TikTok, would transition from potential Chinese government influence to certain American corporate and political influence. The algorithm determining which voices get amplified, which narratives gain traction, and which perspectives get suppressed would be controlled by a consortium of wealthy individuals and investment firms rather than democratically accountable institutions.
The proposed deal allocates roughly 45% control to the Oracle-led consortium, 35% to existing ByteDance investors including General Atlantic, Susquehanna, and Sequoia, and 19.9% to ByteDance itself. On paper, this satisfies the requirement that foreign adversary entities control less than 20%. In practice, it creates a complex web of financial relationships where complete independence becomes difficult to verify.
Republican megadonor Jeff Yass, who owns a major stake in ByteDance through Susquehanna and also holds equity in Trump’s Truth Social parent company, exemplifies the political entanglements embedded in this deal. The mechanism for addressing legitimate security concerns becomes indistinguishable from orchestrating transfer of valuable assets to political allies while maintaining the structural vulnerabilities that prompted intervention in the first place.
Surveillance Capitalism’s Blind Spot
Perhaps the most significant insight from the TikTok controversy is what it reveals about the inadequacy of platform-specific solutions to systemic surveillance problems. While Congress fixates on TikTok’s Chinese ownership, similar data collection practices by American companies receive comparatively little scrutiny. The real national security vulnerability is not ByteDance’s ownership but the absence of comprehensive federal privacy legislation limiting how all companies collect, store, analyze, and monetize personal data.
The U.S. government must obtain court orders backed by evidence to compel American companies to provide user data. By contrast, China’s 2017 National Intelligence Law requires Chinese companies to assist with intelligence gathering upon request. This difference matters, but it addresses only government access to data, not the far broader ecosystem of data brokers, advertisers, and third-party services that buy and sell user information with minimal oversight.
Research from the RAND Corporation highlights a threat receiving insufficient attention: TikTok’s 34 million videos posted daily serve as ideal training material for deepfake-generating artificial intelligence systems. This threat exists regardless of who owns the platform, though concern intensifies when that data flows to foreign adversaries. The Center for Strategic and International Studies documented how TikTok censored content about Hong Kong’s pro-democracy protests, Tiananmen Square, Tibetan independence, and Falun Gong, demonstrating how algorithms shape public discourse without users recognizing manipulation.
Documented incidents reveal the vulnerability of even supposedly isolated systems. ByteDance employees in China improperly accessed data on American journalists investigating internal leaks, leading to four firings including two China-based employees. The incident demonstrated the difficulty of truly isolating data and systems when the parent company retains operational involvement, precisely the vulnerability a licensing arrangement perpetuates.
TikTok generated $10 billion in U.S. revenue in 2024, part of global revenues reaching $23 billion, with advertising revenue projected to hit $33.1 billion in 2025. ByteDance, valued at approximately $300 billion, has seen its international business revenue increase by 60% in the first half of 2024. These financial stakes explain why ByteDance resists genuine divestiture while accepting a licensing arrangement that preserves ongoing technical and financial relationships.
The Path Forward: Beyond Corporate Theater
The Supreme Court unanimously upheld the Protecting Americans from Foreign Adversary Controlled Applications Act on January 17, 2025, rejecting TikTok’s First Amendment challenge and finding that Congress’s national security concerns justified requiring ByteDance to divest by January 19, 2025. Former President Biden could extend that deadline by 90 days if divestiture progress is demonstrated. Trump, who changed his stance on TikTok during the 2024 campaign and urged voters to support him to “save TikTok,” now faces pressure to extend the deadline further as his administration works to finalize the licensing arrangement.
Moolenaar is awaiting a detailed briefing on the deal’s structure, but his preliminary assessment suggests prominent Congressional voices remain unconvinced that licensing the algorithm constitutes genuine divestiture. His warning serves as a reminder that genuine security requires more than creative corporate restructuring: “I just believe you have to have a new algorithm, and I don’t know that you can reprogram.”
The resolution demands solutions transcending the binary choice between banning TikTok or accepting licensing arrangements that preserve the vulnerabilities they claim to address. First, the government must develop genuine technical expertise in auditing and verifying algorithmic independence. Claims that an algorithm has been “retrained” and is “monitored by trusted partners” mean little without transparent, rigorous verification processes detecting hidden backdoors, data exfiltration mechanisms, or subtle influence embedded in machine learning models.
Second, any resolution should catalyze comprehensive federal privacy legislation applying to all platforms, not just those with foreign ownership. Such legislation should limit data collection to what is genuinely necessary, restrict data sharing with third parties, require transparency about algorithmic curation, and establish meaningful penalties for violations. The European Union’s approach, while imperfect, demonstrates that democratic societies can impose constraints on surveillance capitalism without wholesale platform bans.
Third, policymakers must confront the uncomfortable reality that controlling TikTok’s algorithm means controlling what millions see and believe. Whether that control rests with ByteDance, Oracle, or any entity, it represents a concentration of influence over public discourse sitting uneasily with democratic principles. Long-term solutions may require rethinking how social media platforms operate, potentially moving toward more decentralized or user-controlled models distributing algorithmic power rather than concentrating it in corporate hands.
The deadline for finalizing the TikTok deal has been repeatedly extended, most recently to a target of completing the transaction by December 2025 while securing Chinese regulatory approval. The outcome will set a precedent not just for TikTok but for how the United States addresses foreign ownership of digital platforms in an era when data and algorithms have become instruments of geopolitical competition.
The question is whether national security concerns will drive thoughtful reform of surveillance practices and platform governance, or simply provide cover for transferring control of valuable assets to politically connected insiders while leaving fundamental vulnerabilities intact. For 170 million Americans whose attention and data remain at stake, the answer matters far more than who signs the licensing agreements.
Discover more from Being Shivam
Subscribe to get the latest posts sent to your email.