President Donald Trump announced on Sept. 19, 2025, a preliminary agreement for the sale of a majority stake in TikTok from Chinese tech giant ByteDance to a group of U.S. investors following Trump’s negotiation with Chinese leader Xi Jinping.

The deal would create a new U.S.-only version of the app, bringing it into compliance with a law signed by former President Joe Biden on April 23, 2024, and upheld by the Supreme Court on Jan. 17, 2025. Specifics of the deal remain to be hammered out, and left unresolved is the fate of the video sharing app’s core algorithm—and what that means for TikTok’s millions of U.S. users.

The Chinese government has indicated it will not permit ByteDance to sell the algorithm, because it is classified as a controlled technology export, per Chinese law. Meanwhile, U.S. tech industry executives and some lawmakers say compliance with the law requires the algorithm to be under American control. The deal as proposed includes licensing the algorithm so that it remains Chinese intellectual property while the U.S. version of the app continues to use the technology.

TikTok’s For You Page algorithm is widely considered the most important part of the app. As one analyst put it: “Buying TikTok without the algorithm would be like buying a Ferrari without the engine.”

The algorithm’s value lies in its uncanny capacity to anticipate users’ content preferences. Many users claim it knows them better than they know themselves—a sentiment that has evolved into a curious mix of spiritual belief and conspiracy theorizing, as my colleagues and I have documented. Other scholars have similarly noted that users feel more intimately seen and known by TikTok’s algorithm than those powering other popular platforms.

I have studied social media algorithms for nearly a decade, exploring how our relationships with them have evolved as they become increasingly entwined with daily life. As both a social media scholar and TikTok devotee, I want to shed some light on how the algorithm works and how the app might change in the wake of its sale.

How the TikTok algorithm works

In some ways, the TikTok algorithm does not differ significantly from other social media algorithms. At their core, algorithms are merely a series of steps used to accomplish a specific goal. They perform mathematical computations to optimize output in service of that goal.

There are two layers to the TikTok algorithm. First, there is the abstract layer that defines the outcome developers wish to accomplish. An internal document shared with The New York Times specified that TikTok’s algorithm optimizes for four goals: “user value,” “long-term user value,” “creator value” and “platform value.”

But how do you turn these goals into math? What does an abstract concept like “user value” even mean? It’s not practical to ask users whether they value their experience every time they visit the site. Instead, TikTok relies on proxy signals that translate abstract outcomes into quantifiable measures—specifically, likes, comments, shares, follows, time spent on a given video and other user behavior data. These signals then become part of an equation to predict two key concrete outcomes: “retention,” or the likelihood that a user will return to the site, and “time spent” on the app.

The TikTok For You Page algorithm relies on machine learning for predicting retention and time spent. Machine learning is a computational process in which an algorithm learns patterns in a dataset, with little or no human guidance, to produce the best equation to predict an outcome. Through learning patterns, the algorithm determines how much individual data signals matter for coming up with a precise prediction.

A Wall Street Journal investigation found that the amount of time users spend watching each video plays a large role in how the algorithm chooses videos it suggests to users. Using the equation it has generated to predict retention and time spent, the algorithm assigns a score to each video and ranks possible videos that could be shown to the user by this score. The higher the score for an individual user, the more likely the video will appear in their feed.

Of course, content characteristics and other users additionally inform recommendations, and there are other subprocesses folded into the equation. This step is where algorithmic moderation usually comes in. If a video looks like engagement bait or has excessive gore, for example, the content’s score will be penalized.

What’s likely to change for US users

The sale has not been finalized, and what happens to the algorithm is unresolved. However, it’s fairly certain that TikTok will change. I see two key reasons for change.

First, the proposed app’s U.S.-only user population will alter the makeup of the underlying dataset informing algorithmic recommendations on an ongoing basis. As the kinds of content and users come to reflect American cultural preferences, values and behaviors, the algorithm may be slightly different as it “learns” new patterns.

Moreover, not all users will choose to join the new app, especially if it is seen as under the control of Trump’s allies. The current deal reportedly would give an 80% share to U.S. investors, including 50% to new investors Oracle, Silver Lake and Andreessen Horowitz. These investors’ have connections to Trump, and an apparent provision of the deal allows the U.S. government to select one board member. This may result in a user population—and data—reflective of a narrower realm of interests and ideologies.

Second, it’s possible that the majority share owners of the new app will decide to adjust the algorithm, particularly when it comes to content moderation. The new owners may wish to modify TikTok’s Community Guidelines according to their view of acceptable and unacceptable speech.

For example, TikTok’s current Community Guidelines prohibit misinformation and work with independent fact-checkers to assess the accuracy of content. While Meta used to follow a similar approach for Instagram and Facebook, in January 2025 it announced that it would end Meta’s relationships with independent fact-checkers and loosen content restrictions. YouTube has similarly relaxed its content moderation this year.

The bottom line is algorithms are highly sensitive to context. They reflect the interest, values and worldviews of the people who build them, the preferences and behaviors of people whose data informs their models and the legal and economic contexts they operate within.

This means that while it’s difficult to predict exactly what a U.S.-only TikTok will be like, it’s safe to assume it will not be a perfect mirror image of the current app.

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Looking for a compact keyboard with a unique twist? The Keychron Q1 HE (9/10, WIRED Recommends) is currently marked down over $35 on Amazon, and comes with Hall Effect switches, a rare offering that adds a ton of functionality to your keyboard.

Think of a keyboard switch like light switches. They have a fixed point where they activate, and all they can report is whether they’re in one of two states. Hall effect switches are more like light dimmer knobs. They know exactly where they currently are and can report that information back to the computer, which has a number of advantages over traditional keyboard switches.

For starters, you don’t have to settle for a fixed actuation point like you do on most keyboards. You can use the software to set the keys to be super sensitive, or require them to be almost all the way down, or even set keys to send different button presses depending on how hard they’re pressed. If you like to play video games, you can set keys to act like an analog joystick or trigger, letting you easily steer in racing games or walk in RPGs without picking up a controller.

The software has a lot of options, but is well thought out and easy to use, although you do need to plug it in to make changes. It’s QMK-based, but Keychron provides their own web-based launcher to make things even easier, particularly if you’re not well versed in that customization software.

It’s a premium mechanical keyboard throughout. The switches themselves are made by Gateron, and our reviewer noted that they’re exceptionally smooth, thanks to the pre-lubed rails and magnetic sensor. The gasket mount design and full aluminum body work together to provide a soft, deep, typing experience. Unless you’re already using linear switches, you might miss some of the crispness and feedback you get from a tactile or clicky mechanical switch.

While the larger Keychron Q6 HE currently sits at the top of our list of favorite mechanical keyboards, if you don’t need the num pad, you might appreciate the extra desk space you can reclaim with the Q1 HE. They’re otherwise extremely similar boards, and you’ll save $50 in the process.

Manisha Krishnan: So I think it’s kind of important to contextualize this because this is happening at a time where we’re seeing age verification being applied to a range of industries from porn to video games, and companies are going about it in different ways. And while I think there obviously is a lot to be concerned about with young people having unfettered access to ChatGPT, I think these efforts always raise a lot of questions like how is age going to be verified? Where’s that data going to be stored? Even the idea of something inappropriate flagging someone’s parents or the authorities, what is appropriate?

Zoë Schiffer: Right.

Manisha Krishnan: I can definitely see why suicide would be something that you’d want to flag, but maybe there’s other things that the authorities or someone’s parents may not be helpful in every situation. And I guess I’m thinking more about when we go into young people’s exploration of their sexual orientation, just as one example as the culture war topic that’s big right now. So maybe I’m getting too ahead of myself, but those are some of the questions that come to mind.

Zoë Schiffer: That was immediately what came to mind for me, especially because you specifically report on the adult content industry, and I feel like this issue always comes up there where it’s what’s the trade-off between privacy and keeping people safe? But when it comes to young people, it really feels like the privacy conversation goes out the window and regulators are much more inclined to be like, safety comes first, and we might not necessarily care if we’re degrading privacy in some kind of fundamental way.

Manisha Krishnan: Yeah, with the whole porn conversation too, Pornhub obviously has a litany of controversy, but at this point, because they’ve been in trouble so much, they’ve buttoned up a lot of their regulations. And now, in response to the age verification stuff, they’ve removed themselves. So you’re also like, “Am I opening up a vacuum to other maybe more nefarious or irresponsible sites?” There’s always something that will crop up in replacement of something else.

Zoë Schiffer: Well, we’ll wait to see how OpenAI continues to handle this. Staying on the topic of AI for one moment, our next story is about how US tech giants are investing billions of dollars in AI infrastructure in the UK. Our colleague, Natasha Bernal, reported that Microsoft and NVIDIA announced that they’ll be investing up to $45 billion in the form of data centers and AI research. This comes on the heels of another joint venture from NVIDIA, Nscale and OpenAI that’s also aimed at boosting AI infrastructure in the country. Earlier this week, OpenAI CEO, Sam Altman, and NVIDIA CEO, Jensen Huang, traveled with President Trump to the UK during his state visit and then we got a whole bunch of announcements about all of these billion-dollar investments.

Manisha Krishnan: Honestly, one of my first reactions was is this just another form of American tech imperialism spreading and how do the Brits feel about this and what is the underlying motivation for these tech companies to make all these announcements? Do they actually want to invest that much in the UK? That would make sense, but also is it to appease Trump?