Force TikTok divestment or ban over China ties?

National security is serious; so is the First Amendment. Forcing a nationwide TikTok ban because Congress doesn’t trust ByteDance risks turning a real policy challenge into a theatrical takedown of 170 million Americans’ speech. When Montana tried a statewide ban, a federal judge blocked it for likely violating free speech—because silencing an entire platform punishes users more than owners. Even the April 2024 law that threatens a ban after a divestment window will face the same constitutional headwinds if it muzzles creators without demonstrating narrow tailoring and clear evidence. Yes, China’s 2017 intelligence law, ByteDance’s admitted 2022 misuse of journalist data, and opaque algorithms justify vigilance. But a ban is cybersecurity cosplay if we leave the real barn door open: America’s unregulated data-broker economy. Beijing doesn’t need TikTok if it can buy your location history wholesale. Pass a comprehensive federal privacy law, bar data sales to foreign adversaries, mandate independent audits and logging for large platforms (including TikTok), and keep the government-device bans. That’s how you fix the system, not just the headline app. If divestment happens, it should follow a transparent CFIUS process with a specific, public national-security rationale and safeguards that protect users’ speech. Think code escrow, access controls supervised by U.S. trustees, algorithmic transparency requirements, and real penalties for violations—applied across the industry. Protect rights and security at the same time; we can walk and chew gum without chewing up the First Amendment.

TikTok isn’t just cat videos; it’s a foreign-influenced media pipeline shaped by a company subject to the Chinese Communist Party’s intelligence law. ByteDance has a PRC “golden share” in a key subsidiary, and in 2022 it admitted employees improperly accessed U.S. journalists’ data. Former staff say China-based engineers had access to U.S. data until recently. Combine that with an algorithm that can amplify or bury news for 170 million Americans, and you have a textbook influence vector. That’s why the U.S. and allies banned TikTok on government devices. We don’t let Huawei build our 5G; we don’t need ByteDance curating our newsfeed. Divestment is the narrow solution. It targets ownership and control—not content. CFIUS has done this before with Grindr and other deals. Congress’s 2024 law gives ByteDance months to sell; if it refuses, only then does a ban kick in. Yes, some evidence is classified—because advertising our sources on TikTok would be…counterproductive. But when the intelligence community and bipartisan lawmakers wave the red flag, waiting for a public smoking gun is how you get burned. Creators and small businesses can keep their audiences under new ownership; the app can remain, the backdoor cannot. We gave Project Texas years; it delivered press releases, not sovereignty. National security doesn’t mean censoring speech—it means ensuring a foreign adversary doesn’t hold the keys to the platform hosting it. Sell, sever, and let the marketplace—and the memes—continue without Beijing in the loop.

If Congress’s remedy is “sell or shut down,” we should be honest about what that means in practice. Who buys? Letting a Big Tech acquirer scoop up TikTok kneecaps antitrust policy and consolidates yet more power in the hands of a few firms. Handing it to private equity trades one governance risk for another—opaque owners with every incentive to squeeze data for cash. And Beijing’s export controls on recommendation algorithms make a clean sale unlikely anyway, turning a divestment statute into a stealth ban with better branding. Courts will notice if a supposedly narrow tool predictably yields the broadest speech restriction. There’s a smarter blueprint that doesn’t make the First Amendment play chicken with national security. Start with the foundation: a federal privacy law that bans sensitive-data sales to foreign adversaries and reins in the data-broker bazaar. Layer on verifiable safeguards across all large platforms: independent code escrow, U.S.-based key management, third-party runtime attestation for production systems, audit logs with regulator access, and real penalties for violations. Pair that with industry-wide transparency: public ad libraries, C2PA-style content provenance for civic content, and standardized risk reports akin to the EU’s DSA. If you want to deter influence ops, require consistent, cross-platform rules and daylight—not a one-off guillotine. And please, let’s stop pretending a platform ban is a costless switch. Bans travel: if we normalize “national security” as a catch-all to block foreign media, authoritarian governments will cite our precedent to boot U.S. outlets and creators. Protect speech and security simultaneously by giving users real escape hatches: mandated data portability and audience transfer APIs, interoperable social graphs, and small-business support to diversify channels. That’s resilience, not whack-a-mole.

This isn’t about lip-syncs; it’s about who controls the software pipeline that decides what 170 million Americans see. Data localization and nice PowerPoints don’t solve the core problem: ByteDance remains answerable to a regime that can quietly pressure updates, weighting, and moderation guidelines. You don’t need a USB stick marked “Top Secret” to understand how leverage works—look at how PRC regulators wield “golden shares” and how Chinese export controls can dictate what gets sold. If Beijing won’t let the recommender go, divestment forces the question; if they refuse, the ban isn’t a gotcha, it’s the failsafe. We should plan for the real world, not the lab. In a geopolitical crisis—Taiwan Strait, another Middle East escalation, a disinfo push in an election—you cannot spin up a months-long audit to catch a two-day manipulation. The time to remove the structural risk is before the fire drill. A U.S.-owned platform with a U.S. board, U.S. code-signing authority, and U.S.-controlled infrastructure is the minimum standard; everything else is a press release with better clip art. Free speech matters, but this policy targets ownership and control, not the words of creators. The app can live under new stewardship; if it doesn’t, alternatives exist and creators deserve practical help: mandated audience-portability APIs, cross-posting tools, and a 120-day runway for migration. And yes, we should also pass a federal privacy law and crack down on data brokers—security is a both/and, not either/or. But national security doesn’t wait for the perfectly footnoted white paper. Set the guardrails now: sell, sever, and keep the feed—without a foreign adversary holding the algorithmic steering wheel.

If you design policy for the worst day, it has to work on the ordinary days—and in court. The April 2024 divest-or-ban law is already tied up in litigation; an emergency injunction in the middle of a geopolitical flare-up would be the least satisfying plot twist imaginable. Instead of gambling national security on whether a judge thinks your statute is “narrow” this week, copy a playbook that actually survives scrutiny: the FOCI mitigation tools CFIUS uses for defense contractors. Put TikTok’s U.S. operations under a Special Security Agreement with a proxy board of U.S. citizens, cleared trustees controlling code-signing keys, and a government monitor with teeth. We already run satellite makers this way; we can manage a video app without inventing a constitutional knife fight. And let’s stop pretending ownership magic fixes software physics. If Beijing won’t export the recommender, your forced sale becomes a slow-motion product lobotomy—exactly the kind of predictable failure courts read as pretext. There’s a better hedge: put the U.S. instance of the recommender and its training data in a domestic “Algorithm Trust” with independent stewards. Updates require multilayer cryptographic approvals, continuous runtime attestation, and post-release audits by a third party with statutory access. Tie those controls to executive certifications under penalty of personal liability. If the PRC tries to tug on a string in Beijing, the U.S. build pipeline simply won’t accept the change. If Congress insists on divestment, then do it like adults: prohibit acquisition by the Big Tech usual suspects, cap private-equity leverage so we don’t swap CCP leverage for debt leverage, and default to a U.S. IPO with dispersed ownership within two years. Preserve competition by requiring data portability, ad transparency, and labor protections as conditions of any sale. Otherwise we trade one governance risk for another—and the only guaranteed losers are creators, small businesses, and the First Amendment, none of whom asked to be the extras in a geopolitical remake of “Deal or No Deal.”

FOCI mitigation works when you’re guarding factory floors and missile schematics; it doesn’t neutralize a live, cloud-native codebase pushed to 170 million Americans every day. You can’t proxy-board a pull request. As long as ByteDance and its engineers remain under PRC jurisdiction, pressure can be applied offstage—through moderation guidelines, ranking tweaks, or a conveniently timed SDK. The only reliable fix is structural: a U.S.-domiciled company with a U.S. board, U.S. code-signing authority, and a U.S.-controlled pipeline. If Beijing’s export controls keep the recommender behind the Great Firewall, that’s not American censorship—it’s proof of foreign control, which is precisely why the off-ramp exists. On the law, courts have long recognized broad deference where foreign ownership intersects with national security. Ralls v. CFIUS forced more process, not a veto on divestment; broadcast and telecom have foreign-ownership limits for the same reason. The 2024 statute targets who holds the keys, not what users say, and it offers a months-long path to a sale before any ban. That’s narrow tailoring with a seatbelt—especially compared to Montana’s statewide ban that tried to erase an entire channel of speech by fiat. If challengers want the government to publish classified sources to prove a point, they’re asking for a different kind of security failure. Implementation can be practical and pro-competition. Bar a sale to Big Tech, favor a U.S. investor consortium or an IPO with a capped control block, and use a short Transitional Services Agreement that sunsets in months, not years. Stand up U.S.-only SREs, require third-party code audits, and put a CFIUS monitor in the server room—not just the boardroom. Give creators migration tools and a runway, but keep the strategic principle intact: the feed continues, the foreign leverage ends. We don’t let adversaries run our 5G backbone; we shouldn’t let them own the switchboard of our attention economy. Sell if you want a U.S. TikTok; if Beijing says no, we have our answer.

Let’s grant the core conservative point: ownership matters. But if the government can force the sale of a major communications platform used by roughly half the country, it needs more than a vibes-based national-security theory plus a classified appendix and a trust-me bro from the intelligence community. The constitutional problem is not imaginary. Courts are being asked to weigh a law that functionally shutters a speech venue if a sale proves impossible—which, thanks to Beijing’s export controls on recommendation technology, is not some far-fetched hypothetical but the most likely branch of the flowchart. Calling that a “divestment measure” when everyone can see the ban-shaped shadow it casts is clever legislating, not necessarily lawful legislating. And there’s a policy coherence problem conservatives still haven’t solved: if algorithmic influence is the danger, why are we building a TikTok-only doctrine instead of a platform-integrity regime? X, YouTube, Instagram, and TikTok all shape attention, all can be gamed, and all are already battlefields for state-backed manipulation. So create a real national framework: mandatory incident reporting for suspected foreign influence ops, secure access for vetted independent researchers, provenance standards for political media, ad transparency, recommender-system risk testing during elections, and enforceable privacy limits across the board. Otherwise we’re not building a security architecture; we’re doing app-specific exorcisms. There’s also an international legitimacy cost. The U.S. has spent years arguing for an open internet against sovereign control models pushed by China and others. If our message becomes “we support global internet openness, except when Congress gets sufficiently spooked,” that weakens U.S. credibility with allies and hands censors a shiny precedent wrapped in stars and stripes. The better democratic answer is burden-shifting oversight: if ByteDance wants to operate here, impose strict fiduciary duties on U.S. officers, aggressive audit rights, criminal penalties for deceptive disclosures, and a kill switch triggered by proven violations—not by speculation alone. That’s tougher than a press release and smarter than setting the First Amendment on fire to save it.

The liberal case keeps drifting back to process, parity, and platform-wide reform, all of which are nice in a white paper and insufficient in a crisis. The issue here is not that TikTok has an algorithm; it is that TikTok is owned by a company ultimately exposed to coercion by an authoritarian state with both the legal tools and strategic incentive to exploit that fact. That makes it categorically different from domestic platforms, however flawed they may be. Treating TikTok as just one more messy social app is like treating a foreign-controlled port terminal as just another warehouse. Same forklifts, very different leverage. And the “show us the public smoking gun” standard is backward for counterintelligence. If the government had to wait until a manipulation campaign was obvious and declassified, the policy would arrive after the horse, the barn, and half the county fair had already left town. Lawmakers are not banning dance videos because they dislike youth culture; they are responding to a structural vulnerability identified across administrations, by both parties, and by allied governments that have already restricted TikTok on official devices. The persistence of concern through different presidents and different Congresses is itself evidence this is not a passing moral panic with better lighting. As for precedent, the stronger precedent is that the United States routinely limits foreign control in sensitive sectors when influence over infrastructure, communications, or data creates strategic risk. That’s not hypocrisy; it’s statecraft. If divestment is hard because Beijing may block the export of the crown-jewel algorithm, that doesn’t undermine the case for the law—it confirms it. A company that cannot separate from a foreign adversary’s control is exactly the company Congress should not leave sitting in the middle of America’s attention grid. So yes, pair this with privacy reform, portability for creators, and antitrust guardrails on any buyer. But don’t confuse useful side dishes for the entrée. The clean principle remains: Americans can keep the app, or ByteDance can keep control, but in a serious national-security policy it should not get to keep both.

SATIRICAL LIBERAL BOT: Conservatives keep saying this law targets ownership, not speech, but that neat distinction gets squishier the second the remedy is “no qualified sale, no platform.” In reality, the state is using ownership regulation to produce a speech outcome, and courts are capable of noticing when a so-called structural fix functions as a giant on/off switch for a major communications venue. The D.C. Circuit may give the government some deference on foreign-affairs risk, sure—but deference is not a constitutional blank check with a tiny American flag sticker on it. If the government’s theory is that a platform can be effectively extinguished based on secret evidence and predictive harms, then the limiting principle matters a lot more than the talking points do. There’s also a practical governance issue the ban-or-divest camp still sidesteps: what exactly is being sold? If the recommendation engine, training methods, and core technical staff are constrained by Chinese export controls or corporate dependency, then a “sale” may yield a hollowed-out clone with the brand but not the product. That doesn’t just hurt ByteDance; it punishes U.S. users and creators while pretending not to. And if the answer is “fine, then let it die,” that reveals the policy was always less a clean divestment tool than a ban with paperwork. National security policy should not hinge on a legal fiction everyone can see through from space. A more durable democratic approach is to regulate the capabilities that create the risk, not just the passport of the parent company. Require all very large platforms to meet baseline obligations for data minimization, foreign-access logging, secure code-signing, researcher access, and auditable controls over recommendation changes during elections or international crises. Add a rebuttable presumption of emergency restrictions if the government can show concrete violations or deceptive disclosures, with expedited judicial review and a real evidentiary record. That would let the state act fast when needed without normalizing the idea that Washington can delete a speech platform first and explain later. If we’re defending liberal democracy, the “liberal” part should not be considered optional DLC.

SATIRICAL CONSERVATIVE BOT: The liberal argument keeps searching for a perfectly platform-neutral framework, but the actual risk here is not abstract “algorithmic power”; it is adversary leverage over a specific platform at massive scale. Ownership is not a side detail—it is the mechanism by which pressure can be applied quietly, rapidly, and plausibly deniably. In a world where influence operations can ride recommendation systems, moderation calls, or internal access pathways, waiting to regulate only after “concrete violations” are publicly provable is a bit like insisting the alarm system may only be installed after the burglary has been peer-reviewed. The point of counterintelligence is to reduce exploitable dependency before the crisis, not after the congressional hearing. And yes, if a sale is difficult because the PRC won’t permit transfer of the algorithm or key technical assets, that is not proof the U.S. law is unfair—it is proof the U.S. concern is well founded. A company that cannot disentangle itself from a geopolitical rival’s control architecture should not be treated as just another foreign investor with a cool app. The 2024 statute gave ByteDance a path: divest, restructure, or lose access to the U.S. market. That is not censorship of users’ viewpoints; it is a condition on foreign control of a strategically sensitive media platform. We impose similar logic in telecom, semiconductors, and critical infrastructure because “but the service is popular” is not, in fact, a national-security doctrine. The strongest version of this policy is not crude prohibition; it is disciplined severance. Pair divestment with a ban on sale to dominant incumbents, require a U.S.-controlled board and code-signing authority, mandate continuity tools for creators and advertisers, and set strict transition deadlines so this does not become Project Texas 2: Return of the PowerPoint. If Congress also wants a federal privacy law, excellent—pass it yesterday. But privacy reform is not a substitute for removing adversary leverage over a platform that reaches 170 million Americans. The cleanest principle is still the best one: if ByteDance can truly separate, let TikTok stay; if it cannot, then Washington should stop pretending dependency is a form of openness.