TikTok, the widely popular short-form video sharing app, is in the process of overhauling its recommendation algorithm for its 170 million U.S. users, Reuters has reported.
This effort aims to develop a version of the algorithm that operates independently of its Chinese parent company, ByteDance, to address concerns from American lawmakers who have been advocating for a ban or forced sale of TikTok’s U.S. operations.
The initiative to separate the source code was initiated by ByteDance late last year, prior to the congressional bill that demands the sale of TikTok’s U.S. assets gaining significant traction. This bill, which was signed into law in April, mandates a divestiture or ban of the app by January 19, 2025.
Tekedia Mini-MBA edition 15 (Sept 9 – Dec 7, 2024) has started registrations; register today for early bird discounts.
Tekedia AI in Business Masterclass opens registrations here.
Join Tekedia Capital Syndicate and invest in Africa’s finest startups here.
Sources with direct knowledge of the efforts, who requested anonymity, disclosed that ByteDance and TikTok engineers are working meticulously to split millions of lines of code. The goal is to establish a code base that is independent of ByteDance’s Chinese operations and remove any links to Chinese user data.
The separation of the code base could potentially facilitate the divestiture of TikTok’s U.S. assets, although no current plans for such a sale have been announced. TikTok has previously stated that selling its U.S. assets is not feasible due to commercial, technological, and legal challenges. This stance was reiterated in a federal lawsuit filed by TikTok and ByteDance in May, seeking to block the law that mandates the sale or ban of the app.
A U.S. appeals court has set a fast-track schedule to address these legal challenges.
The Complexity of the Task
The task of splitting the recommendation algorithm is described by sources as a complex and tedious process that could take over a year to complete. Engineers are tasked with reviewing each line of code to determine its suitability for the new, independent code base. This process involves sifting through the algorithm, originally developed by ByteDance engineers in China and customized for global markets, including the U.S.
The U.S. government, including President Joe Biden and supporters of the new law, argue that TikTok gives Beijing undue access to vast amounts of data, which could be used for spying or influencing American users. This has led to bipartisan political pressure on TikTok to demonstrate its operational independence from ByteDance.
In response to these concerns, TikTok had previously attempted to silo off U.S. user data through a project known as Project Texas, which failed to satisfy U.S. regulators and lawmakers. Now, the company is escalating its efforts to show its U.S. operations are independent, considering measures such as open-sourcing parts of its algorithm to demonstrate transparency.
One of the significant risks of this separation is that TikTok U.S. might not maintain the same level of performance without the engineering support from ByteDance’s team in Beijing. The recommendation engine’s effectiveness in engaging users is a key factor in TikTok’s popularity, and replicating this success independently could be challenging.
The legal and political battles have already had financial repercussions for TikTok. Any significant changes or disruptions to TikTok’s algorithm could impact its market performance and user engagement. But a sale of the app, including its algorithms, is highly unlikely. In 2020, the Chinese government added content recommendation algorithms to its export-control list, meaning that any divestiture or sale of TikTok’s algorithm must undergo its administrative licensing procedures.