Should the U.S. force a TikTok divestiture or ban the app?

Let’s not hand the government a giant “off” switch for the internet because a committee hearing got spooky. Yes, Huawei-grade concerns about Chinese law and data access are real. But banning a platform—or forcing a shotgun sale under a political deadline—sets a precedent that the government can decide which megaphone the public gets to use. We’ve already watched courts swat down the 2020 TikTok ban attempt and Montana’s statewide ban as First Amendment overreach, and TikTok’s 2024 lawsuit argues the new divest-or-ban law repeats the same sins. If the goal is security, we need platform-agnostic rules that apply to everyone, not one-off crusades that look like content policing in a trench coat. The right move is to fix the actual leak: America’s data-broker free-for-all. Right now Beijing can just buy U.S. location and ad-tech data from brokers while we’re busy performatively unplugging a single app. Pass a comprehensive federal privacy law, ban selling sensitive data to foreign adversaries, mandate transparency on recommendation algorithms, and require real third-party audits—measures that cover TikTok, Meta, X, the whole gang. Project Texas-style data localization and code escrow can be made real with independent monitors and teeth; if TikTok (or any platform) flubs it, impose fines, structural remedies, or a targeted, evidence-based CFIUS order. Free expression isn’t a dessert you get if you clean your cybersecurity plate; it’s the main course. Millions of Americans—creators, small businesses, and communities that don’t get prime-time TV slots—use TikTok for speech and commerce. If the answer to “foreign leverage” is “delete the app,” we’ve learned nothing from past overreaches. Use scalpel laws that actually protect data and curb covert state influence. If divestiture is ever necessary, do it through a transparent process with verifiable technical separation—not a legislative countdown clock that doubles as a kill switch.

A platform controlled by a company subject to Beijing’s National Intelligence Law isn’t just a dance app; it’s an operational risk. ByteDance doesn’t need to hand over raw phone numbers to matter—nudging an algorithm to dial down protest footage or elevate state narratives is the subtlety play. Project Texas was a PR tour, not an air gap; we don’t outsource our information environment to an adversary and call it “innovation.” Congress finally passed a divest-or-ban law in 2024 because CFIUS negotiations stalled for years. India didn’t faint at the thought of a ban; it made a call. Sovereignty sometimes looks like saying “no.” This isn’t censorship; it’s ownership hygiene. We already restrict Huawei, ZTE, and Kaspersky without melting the First Amendment. The statute targets control, not content: TikTok can keep humming if ByteDance sells to an American or allied buyer under strict compliance. Users aren’t speechless without one app—newsflash, the First Amendment doesn’t guarantee you the For You Page. Meanwhile, courts have long upheld limits when national security and foreign adversary control are at stake; we regulate ports and power grids, and the attention economy is no less strategic than steel. Divestiture is the off-ramp. If Beijing blocks a sale, they’ve answered the security question for us. Set a real deadline, enforce it, and pair it with two upgrades: a federal data-broker crackdown that stops our info from being vacuumed up and sold abroad, and app-store security baselines that flag risky SDKs and permissions. Perfect security isn’t on the menu, but “CCP-adjacent control of a dominant U.S. media platform” doesn’t have to be either. TikTok can stay—just not as a branch office of the Politburo.

If the plan is divest-or-delete, be honest: you’re building a Rube Goldberg kill switch. Beijing already slapped export controls on recommendation algorithms, and China’s regulators would have to approve a sale of TikTok’s core tech. That’s not a bug; it’s how you turn a “divestiture” into a de facto ban with plausible deniability. Courts apply heightened scrutiny when the government kneecaps a communications platform, and an ownership test wrapped around a countdown clock is still a speech burden—especially when millions use TikTok Shop and live-streams for livelihood and civic organizing. If you want to prove narrow tailoring, you can’t skip directly to the most speech-restrictive remedy because CFIUS talks got tedious. There’s a grown‑up alternative that doesn’t cosplay as cyber braveheart: mandate platform-agnostic systemic risk audits like the EU’s DSA (with real fines), require provenance tagging for political content, and impose independent, on‑prem code review with board‑level compliance obligations—think special master for algorithms, not vibes. Grindr’s forced sale happened through CFIUS with a detailed record tied to sensitive data; you can do that case-by-case when evidence demands it. If ByteDance fails ring‑fencing or audit benchmarks, you escalate. That’s how you build a record that survives the D.C. Circuit without turning “foreign adversary” into a magic word for “pull it from the app stores.” Also, let’s not pretend vaporizing a major competitor is neutral. Wiping out TikTok would hand Meta and Google a congratulatory fruit basket and concentrate the U.S. attention markets even more. We didn’t ban Facebook after Russia’s 2016 antics; we demanded transparency and enforcement. Apply the same rules here—equal, verifiable, and boringly effective. Security should look like plumbing, not a magic trick where you smash one faucet and hope the pipes behave.

The problem isn’t theoretical. ByteDance admitted in 2022 that employees improperly accessed data on U.S. journalists to hunt down leaks. That’s not a teenage dance scandal; that’s an ownership-and-control scandal. Pair that with China’s National Intelligence Law and the export controls Beijing slapped on recommendation algorithms, and you have a company structurally unable to guarantee independence from state leverage. The algorithm isn’t just code; it’s a programmable editor-in-chief. If you wouldn’t let a foreign ministry staff the newsroom, don’t let it own the feed. The 2024 statute is narrow and ordinary by national-security standards. It targets control, not content; it offers an off‑ramp via divestiture; and it’s reviewable in court. We’ve been here: CFIUS forced the sale of Grindr to remediate data exposure, the U.S. restricted Kaspersky because of Russian leverage, and the FCC has long limited foreign ownership in broadcast licensees. Regulating who controls critical distribution infrastructure is not First Amendment heresy; it’s Guardrails 101. If China refuses to approve a sale of the algorithm, that only confirms the source of risk and clarifies why a ban becomes necessary. And yes, pair divestiture with broader fixes—data-broker rules, supply‑chain scrutiny, and app‑store baseline security. But don’t confuse “we should do more” with “we should do nothing.” A platform with 170 million U.S. users is a strategic asset, not a startup. We don’t outsource ports, power grids, or broadcast towers to adversary‑subject companies; the attention economy belongs on the same list. TikTok can absolutely stay—but under ownership that answers to U.S. law, not to a Politburo veto.

If you think divest-or-ban is clean constitutional living, try it against Packingham’s backdrop where the Court called social media the modern public square. Compelled ownership change under a political countdown is still a burden on speech—even if you insist it’s just corporate housekeeping. And the list of “foreign adversary-controlled platforms” won’t live in a glass case labeled Break Only for ByteDance; it’s a precedent a future Congress can aim at whatever platform is politically inconvenient. Add the Fifth Amendment gray zone around compelled sales and you’ve built an invite to years of litigation while creators and small businesses twist in the wind. There’s a less cosplay-more-compliance route. Do a binding FTC/DOJ consent decree with a court-appointed technical trustee holding the U.S. keys, requiring immutable audit logs of access to U.S. data, and SEC-style “material algorithm change” disclosures with independent verification. Mandate provenance tagging for political content and rapid incident reporting to a neutral regulator—think airline safety board for feeds—plus a government‑devices ban (already common) and critical‑infrastructure network restrictions without gagging the public sphere. If the trustee finds covert manipulation or unauthorized access, you escalate through CFIUS with a fact-rich record tailored to the breach, not a one‑size legislative guillotine. And yes, protect the ecosystem we actually have. Creators shouldn’t be collateral damage in a geopolitical arm wrestle. Bake in portability and interoperability requirements so audiences and shops aren’t captive, and use platform‑agnostic systemic risk audits with real fines to make every giant—from TikTok to Meta—prove they can keep state influence out. Security that looks like plumbing beats a dramatic kill switch that doubles as an antitrust love letter to incumbent U.S. attention monopolies.

Capability plus compulsion equals risk, and TikTok has both. FBI Director Wray has warned about the leverage Chinese law creates over companies like ByteDance, and ByteDance already admitted employees improperly accessed U.S. journalists’ data to hunt leaks. You don’t need a smoking-gun feed manipulation on C-SPAN to act; an algorithm is a programmable editor, and Beijing’s export controls on it are the tell. Ownership hygiene isn’t performative—it’s how you prevent an adversary’s veto from sitting inside a dominant U.S. distribution channel. The First Amendment protects people’s speech; it doesn’t entitle foreign adversary–subject firms to control the pipes. We’ve long limited foreign control of broadcast infrastructure under Section 310(b) and barred risky vendors from sensitive networks; federal and many state governments already ban TikTok on official devices without the Republic imploding. The 2024 law is narrow: it targets control, offers an off‑ramp via sale, and is reviewable in court. If Beijing refuses to approve a sale of the algorithm, that’s not our overreach—that’s confirmation of state leverage, and the ban becomes the necessary backstop. Divestiture keeps competition and creators intact: TikTok can remain, but under ownership that answers to U.S. law. Set clear conditions—U.S.‑based board with security clearances for key infosec roles, escrowed U.S. keys, and an independent monitor with inspection rights—and pair it with a data‑broker crackdown and app‑store supply‑chain vetting so we’re not securing one door while leaving the windows open. Call it boring if you want, but it beats pretending a 170‑million‑user platform owned under a Politburo veto is just another dance app. Turn off the gas before the building blows; then let the music continue safely.

If you want to make the open internet look like a patchwork of tollbooths, congratulations—divest‑or‑ban is how you teach every government to grab the steering wheel of speech platforms they don’t like. Today it’s ByteDance; tomorrow Brazil, Turkey, or a post‑election Poland tells YouTube or X: sell to my buddy or get out. That’s not a paranoid fanfic; it’s how reciprocity and precedent work. We can’t sermonize against China’s Great Firewall while beta‑testing our own platform‑specific takings that invite bill‑of‑attainder claims (yes, TikTok’s 2024 lawsuit says the quiet part out loud) and turn Congress into a content‑adjacent landlord. Murthy v. Missouri already showed the Court’s queasiness with government pressure on the feeds; now we’re proposing a statutory choke chain for one of the biggest. And the “just sell it” fantasy runs into physics. China’s export controls on recommendation algorithms mean a clean sale likely yields a TikTok‑shaped shell without the thing that makes it TikTok—functionally a slow‑motion ban with extra paperwork. Meanwhile, creators get pulverized in the transition, and the market consolidates neatly into Meta and Google’s welcoming arms. We don’t call it national security when the fix doubles as an antitrust valentine to incumbents. There’s a grown‑up route with teeth that doesn’t cosplay as a guillotine. Stand up a Foreign Influence Platform Duties regime: real‑time distribution transparency APIs; immutable, third‑party‑verifiable access logs; whistleblower bounties for covert manipulation; and client‑side integrity attestation using secure enclaves so “oops, a rogue engineer” stops being a plot twist. Add data sanctions that actually bite—ban sensitive U.S. data flows to PRC‑linked buyers and enforce cloud/colo blocklists—plus a diaspora‑protection unit that hunts coordinated influence ops across platforms. Make social‑graph and storefront portability mandatory (ActivityPub‑style), so no audience is taken hostage by any app’s feed. Then coordinate with the EU/UK/Canada on joint audits under the DSA‑ish model already breathing down TikTok’s neck. Security should look like a circuit breaker, not a magic trick where we smash one platform and hope the pipes fix themselves.

We don’t let a foreign adversary own a satellite uplink and promise to ‘audit harder’; we take the uplink out of their hands. Same principle here. Audits and special masters are catnip for PowerPoint, but model weights update overnight and shadow features don’t leave thank‑you notes. Ask Kaspersky how well “just trust our controls” aged, or Volkswagen’s emissions monitors how easy it is to game an audit regime. A sovereign that can write your patch notes (see: Beijing’s algorithm export veto) doesn’t get to co‑own the programmable front page for 170 million Americans. This isn’t some bespoke vendetta. Properly drafted, the rule is general and forward‑looking: no adversary‑subject control over large‑scale distribution platforms, with a size threshold, a divestiture off‑ramp, and court review. That’s not a bill of attainder; it’s ownership hygiene with national‑security receipts, and it fits comfortably alongside Section 310(b) foreign‑ownership limits and our Huawei/Kaspersky precedents. WTO hand‑wringing? The security exception exists for a reason, and China’s own blanket bans on U.S. platforms make reciprocity arguments… let’s say unpersuasive. Implementation is not a vibe, it’s a plan: a 12–18‑month window for sale; a clean‑room rebuild of the recommender using U.S.‑collected data; a transition‑services agreement to keep creator accounts, ad campaigns, and TikTok Shop intact; and a continuity fund financed by civil penalties plus a temporary platform levy to backstop small businesses during the handoff. If Beijing blocks the algorithm export, we flip to TikTok‑US—same UX, new recommender—under an independent monitor with inspection rights, cleared security leadership, and escrowed keys. Pair it with portability mandates so creators keep their audiences either way and a real data‑broker export ban so we’re not locking the front door while selling the house plans on eBay. Call that boring if you want. It’s also how you prevent a Politburo veto from squatting inside our information supply chain during an election cycle. We can keep the app and the creators; we just can’t keep the control structure that treats an American feed like a subsidiary of Chinese law. Divest if possible, ban if necessary, and stop pretending a ‘third‑party algorithm whisperer’ beats a sovereign with root access.

Calling this “ownership hygiene” doesn’t change that Congress drafted a law with a silhouette that looks suspiciously like one app. Courts aren’t blind to motive: if you burden the dominant forum where millions speak, you need a record that survives more than a press conference. Under Turner and O’Brien, the government must show real evidence and narrow tailoring; under Packingham’s spirit, a divest‑on‑a‑deadline that predictably collapses into a ban looks less like plumbing and more like prior restraint in business casual. Add the Berman Amendment’s long‑standing carve‑out for informational materials and you’ve got a judge’s migraine, not a clean constitutional bill of health. Even if you clear the legal bar, the implementation plan reads like a sprint triathlon in a hurricane. A “clean‑room” algorithm rebuild in 12 months while preserving trust‑and‑safety nuance, creator monetization, live‑commerce integrity, and elections‑period integrity? TikTok’s moderation heuristics, classifier thresholds, and enforcement playbooks are not IKEA furniture with a universal Allen key. The EU just forced TikTok to suspend its “Lite” rewards program and opened formal DSA proceedings—evidence that even mature teams struggle under rapid changes. Now layer an injunction ping‑pong through the D.C. Circuit: creators don’t get paid in certioraris, advertisers freeze budgets, and foreign investors read the tea leaves that U.S. rules can turn into platform‑specific forced sales when the wind shifts. The adult table path is dull and effective: a federal data‑broker licensing regime and a bright‑line ban on selling sensitive U.S. data to PRC‑linked buyers; outbound investment controls for data and AI codified beyond the 2023 EO; DSA‑style systemic‑risk audits with joint EU/UK/Canada teams; and a Public Algorithm Trustee with subpoena power, immutable access logs, and whistleblower bounties for covert manipulation. Add portability mandates for social graphs, storefronts, and ad campaigns so no audience is hostage. If these controls surface state leverage or undisclosed tampering, you walk into CFIUS with a fact pattern and escalate proportionately. That’s deterrence that survives court and doesn’t teach every government to run a “sell to my friend or else” special on platforms they dislike.

We’re not yanking poetry off the shelves; we’re changing the lock on the building. The 2024 statute regulates control, not content, and that matters under O’Brien and Turner. The record isn’t vibes: ByteDance admitted employees misused U.S. journalists’ data; Beijing’s algorithm export controls are an on‑the‑nose tell about state leverage; and the National Intelligence Law is a standing order, not a rumor. The Berman Amendment protects the flow of information, not a foreign adversary’s right to own the distribution channel. Let the D.C. Circuit perform its scrub—this is conduct regulation with a national‑security through‑line courts have blessed in broadcast, telecom, and software precedents. Operationally, we’ve done harder things on tighter timelines. The U.S. executed the 5G “rip and replace,” repacked spectrum, and is onshoring advanced fabs with supply‑chain fire drills. A 12–18‑month divestiture window with a continuity fund, audience and storefront portability, and a U.S.‑based trust‑and‑safety rebuild is attainable—especially compared to the forever‑audit regime that assumes a subpoena can outmuscle a Politburo veto. Audits find problems after the fact; ownership determines who can create them in the first place. If Beijing refuses to let the algorithm travel, that’s the clean signal that a ban is the necessary backstop, not a tantrum. This isn’t a bespoke hit; it’s a rule of general application: no adversary‑subject control of large‑scale distribution platforms, period. China already blocks most U.S. platforms; the EU is hauling VLOPs into DSA compliance and just kneecapped TikTok Lite—so spare me the reciprocity sermon. Pair divestiture with real upgrades: codify outbound‑investment restrictions for sensitive tech and data, pass a data‑broker export ban with criminal teeth, and tighten app‑store vetting for risky SDKs. Toss in FISA 702 reforms to prove we can police ourselves while denying adversaries root access to our feeds. Keep TikTok’s product, creators, and competition; evict the legal backdoor. Divest if Beijing allows it, ban if they won’t—and stop pretending a quarterly audit can outvote a security service with a pen on your org chart.