Divest or ban TikTok? Security vs free speech

Ah yes, nothing says "defend democracy" like Congress deciding which app you’re allowed to be funny on. We’ve spent 20 years failing to pass a basic federal privacy law, but now it’s “national security emergency: teenagers are lip‑syncing.” If you actually care about security, start where the risk lives: the data brokerage free‑for‑all that lets anyone—including Beijing—buy Americans’ location, health, and ad‑ID trails for the price of a nice lunch. A TikTok divest‑or‑ban is whack‑a‑mole with a better press release. The First Amendment doesn’t come with a “except when a platform annoys us” clause. Federal courts already smacked down Montana’s statewide ban and earlier Trump‑era attempts under IEEPA because banning a platform is a blunt instrument that burdens speech. Singling out one app because of ownership—rather than regulating the conduct (data handling, algorithmic transparency, foreign state influence)—is a great way to flunk strict scrutiny. If the government can kill a platform by name today, it can kill a newsroom by pretext tomorrow. That’s not national security; that’s a content pipeline with a government off‑switch. Do the grown‑up thing: a platform‑agnostic privacy law (ban selling sensitive data to brokers, especially to foreign adversaries), mandatory audits and code escrow for high‑risk platforms, robust FARA enforcement and clear labels for state‑linked content, an independent oversight regime for recommender systems, and a bright‑line rule against government influence on any platform’s feed. Keep it narrow, constitutional, and effective. Because if the plan is “ban first, legislate later,” all we’re securing is our reputation for tech policy cosplay.

This isn’t about your dance video; it’s about Beijing doing a duet with your data. China’s National Intelligence Law compels companies to secretly assist its security services—no court warrants, no cute transparency report. ByteDance is a China‑headquartered parent, and we’ve already seen the tells: leaked audio where TikTok staff said “everything is seen in China,” ByteDance employees caught snooping on U.S. journalists’ location data, and repeated warnings from the FBI and the intelligence community that the CCP can pressure owners and tune an algorithm more subtly than a Stradivarius. Project Texas? Great branding, modest substance—Oracle hosting doesn’t nullify corporate control. Divest‑or‑ban isn’t a censorship spree; it’s a supply‑chain fix for the attention economy. We already block Huawei from 5G and Kaspersky from federal systems because infrastructure matters. TikTok’s algorithm is cultural infrastructure—if the CCP can steer or suppress narratives at scale during a crisis or an election, that’s a national security vulnerability, not a vibes problem. Congress (with bipartisan votes) and the White House moved to require a sale on a reasonable timeline; if ByteDance won’t sever control, that’s their business decision, not America silencing memes. And spare me the “end of small creators” dirge. Reels and Shorts exist; audiences follow talent across platforms faster than you can say “link in bio.” A forced sale keeps the content, the creators, and the code—just not the umbilical cord to Zhongnanhai. We’re not banning speech; we’re banning ownership by a company legally obligated to obey an adversary government. If your landlord is the CCP, the fix is simple: get a new landlord, or find a new apartment.

National security speed‑run: Congress stapled a “sell or die” rule to a foreign‑aid bill and declared it constitutional because the vibes said so. The D.C. Circuit suits aren’t about teen dances; they’re about whether the government can put a kill‑switch on a named speech conduit and call it commerce. Face it: if Beijing won’t license the algorithm for export—as it’s signaled before—you didn’t design a divestiture; you engineered a de facto ban, then tried to launder it through corporate paperwork. Narrow tailoring this is not. And spare me the “it’s just ownership” fig leaf. The algorithm is expression in code; forcing a sale of the printing press while knowing the ink won’t transfer is still a prior restraint with extra steps. We tried platform‑by‑platform morality plays and the courts swatted them; writing a bespoke statute to hit one app doesn’t magically solve the First Amendment problem of targeting a specific megaphone used by 170 million Americans. Also, when the United States starts copy‑pasting China’s playbook—naming foreign apps and walling them off—we torpedo our credibility when we tell the rest of the world to keep the internet open. Good luck scolding India, Nigeria, or the EU next time they fancy a ban on a U.S. platform. If the risk is foreign state leverage and data exploitation, regulate the risk—universally. Ban the sale of sensitive data to brokers, full stop. Impose platform‑agnostic security audits with code escrow and real‑time access logs for any high‑risk, high‑reach platform, domestic or foreign. Lock government devices and critical‑infrastructure networks down, enforce FARA and clear provenance labels for state‑linked content, and mandate independent testing of recommender systems across the entire market. You want leverage on Beijing? Use trade, export controls, and diplomacy, not a domestic speech sledgehammer that mostly punishes American users and small businesses while bad actors keep buying the same data from brokers with a corporate credit card.

Let’s retire the “but my memes” defense. Congress didn’t criminalize lip‑syncing; it required that a platform with CCP leash laws on the owner either change owners or exit the U.S. market. That’s Foreign Commerce 101, not a content review board. We’ve done this play before with CFIUS—see Grindr, PatientsLikeMe, StayNTouch—because when an adversary state can compel secret cooperation, corporate “trust us” is not a control, it’s a punchline. And yes, Beijing’s export controls on TikTok’s algorithm are the tell: if the PRC claims the recommendation engine is a strategic asset, maybe we should treat it like one. This isn’t just about data; it’s about cognitive infrastructure. The feed is a dial, not a mirror, and in a crisis that dial can nudge millions. We’ve already seen ByteDance staff misuse data to track U.S. journalists, leaked audio about China access, EU probes into TikTok’s risks for minors and addictive design, and repeated IC warnings that influence ops don’t announce themselves with a watermark. “Project Texas” is window dressing if corporate control stays in Beijing’s blast radius. Huawei taught us you don’t leave adversary‑controlled gear in your backbone; TikTok is the culture backbone—same logic, slicker UI. The First Amendment isn’t diplomatic armor for a foreign government’s leverage. Speech remains free across Reels, Shorts, Twitch, podcasts, Substack, carrier pigeons—you name it. A forced sale preserves the content, creators, and audience while cutting the umbilical cord to Zhongnanhai. If China vetoes the sale, that’s Beijing choosing the ban, not Washington censoring cat videos. And on enforcement: perfect isn’t required. You reduce distribution via app stores and CDNs and you throttle scale—the same way we treat Kaspersky on federal systems. The liberal wishlist of universal privacy law is fine policy homework; it’s not a live fire response. Do the structural fix now; keep debating the term paper later.

Congress basically duct‑taped an export‑control fantasy to a foreign‑aid bill and called it cyber‑patriotism. Cute. But in court you don’t get points for the trailer; you need the receipts. Thus far, the government’s “trust me, it’s classified” routine is doing burpees around the actual First Amendment problem: you’re throttling a uniquely large megaphone with a statute tailored so snugly it still has ByteDance’s name tag on it. And spare me the theater about enforceability—if you want this to bite, you either criminalize workarounds like VPNs (hello, collateral damage) or stage a vibes‑only blockade through app stores and CDNs that savvy teens will treat like a Sudoku. Either way, you’re normalizing a government switch for speech distribution. We’ve seen this movie; it screens poorly in constitutional litigation. Also, your ownership test is a Rorschach blot with subpoenas. “Foreign adversary control” sounds tidy until a U.S. startup takes Saudi money, or a pension fund with a Chinese LP wanders onto a cap table and suddenly we’re speed‑running industrial policy via push notifications. We have a tool for nuanced risk—CFIUS—where you impose auditable, enforceable security agreements, not platform execution by statute. If telecom can live with foreign ownership caps applied sector‑wide, then tech can live with sector‑wide, content‑neutral duties: audits, access logging, code escrow, and penalties that bite when any firm—domestic or foreign—lies, leaks, or launders influence. If the nightmare is subtle manipulation and covert data exfil, regulate the levers everywhere, not the logo. Try this: ban government agencies from buying location or ad‑ID data without a warrant, impose a hard embargo on broker sales to entities tied to foreign intel services, mandate public‑interest APIs so independent labs can live‑audit feed behavior in real time, and require feed choice and algorithmic portability so no single dial owns your attention. Add criminal penalties for undisclosed foreign influence buys across all platforms and election‑period mirror archives of major feeds for post‑hoc review. You want to make Beijing miserable? Dry up the data market and light up the black boxes—no need to bulldoze the stage to fix the wiring.

ByteDance doesn’t live under Vermont LLC law; it lives under China’s upgraded Counter‑Espionage and National Intelligence Laws—comply, secretly, or else. That’s not conspiracy; that’s statute. There’s a party committee inside ByteDance’s China entity, and Beijing’s export controls literally classify TikTok’s recommendation engine as a strategic asset. We just watched TikTok prove its mobilization power by blasting users to melt congressional phone lines—impressive civics lesson, also a demo of what an owner with PRC obligations could do during, say, a Taiwan crisis or a contested election. The whole point of influence ops is you don’t get a smoking gun press release; you get a feed that leans five degrees off center when it matters. Legally, this is familiar terrain. We’ve long limited foreign control over broadcast and telecom infrastructure because structure shapes speech distribution. Turner Broadcasting blessed structural rules with incidental speech effects; Bluman upheld limits on foreign nationals’ participation in our political sphere. The 2024 law targets adversary‑controlled, high‑reach apps via criteria and process, offers judicial review, and gives a surgical remedy: sell to a trustworthy owner or leave the market. That’s not a prior restraint on content; it’s a corporate‑control firewall. If Beijing vetoes the sale, that’s not Washington muzzling memes—that’s the PRC choosing control over American access. And no, a privacy law won’t fix a geopolitical rootkit. Closing brokers is fine; it doesn’t neutralize an algorithm whose owner can be leaned on by a rival state. We tried “Project Texas”; it was theater with better branding. The clean cut is corporate separation—new owner, new fiduciary duty, same creators, same audience. Enforcement doesn’t require perfect: you choke distribution through app stores, CDNs, ad networks, and payment rails, and the scale collapses to rounding error. If creators are as good as they say, they’ll carry their audience to Reels, Shorts, or to a U.S.‑owned TikTok after divestiture. We already keep Kaspersky off federal systems and Huawei out of our backbone; this is the cultural backbone. Treat it like infrastructure, not open mic night.

We’ve now reached the “designate first, rationalize later” phase of governance where a statute with a choose‑your‑own‑adversary clause lets the Executive label a platform radioactive with definitions of “control” stretchy enough to double as a yoga band. That’s not just bad drafting; it’s a First Amendment problem wrapped in a foreign‑commerce costume. Call it what you want—criteria‑based, procedurally seasoned—the practical effect is a bespoke cudgel aimed at a single, massively used channel of speech with a fog‑machine standard that can be repurposed the next time a politician decides a platform’s vibe is “insufficiently deferential.” When Congress makes corporate surgery a precondition for staying in the discourse, you’re flirting with attainder logic and unconstitutional conditions, not cool‑headed risk management. And let’s stop pretending “just go to Reels” is a grown‑up transition plan. Network effects aren’t carry‑on luggage; TikTok Shop alone is a supply chain for micro‑brands that don’t have a Meta ad war chest. India’s ban didn’t make influence risk vanish; it reshuffled it to domestic players with better political Rolodexes. Here, antitrust takes Big Tech off the shortlist, Beijing won’t export the secret sauce, and you’re left with “TikTok, But Make It Zombie”—same logo, worse recommender, creators holding the bag. That’s not a remedy; that’s a vibes tax on small businesses and communities who built an audience where the discovery actually works. If the real fear is covert leverage, then build guardrails that bite everywhere leverage lives. Put TikTok (and any high‑reach app) under a court‑supervised consent decree: tamper‑evident logging from build to prod, hardware‑rooted attestations streamed to a U.S. trustee, cryptographic fingerprints of model updates with public integrity proofs, and criminal penalties for undisclosed “heating” or state‑directed boosts. Pair it with broker licensing and a hard export quarantine on ad‑ID/location data to foreign intel cutouts, and mandate cross‑platform feed‑choice and data portability so no single dial owns the narrative. You want resilience? Don’t bulldoze the stage—install tripwires, floodlights, and a fire marshal who can actually write tickets.

This isn’t a vibes‑based purge; it’s a structural firewall with footnotes. The law doesn’t name‑and‑shame one company; it sets criteria, notice, and review for apps under adversary control—a category Congress is allowed to police in foreign commerce and national security. Courts have long tolerated structural limits that incidentally touch speech distribution (see broadcast ownership caps) and deferred to Congress when foreign influence meets domestic channels. Corporations don’t get a constitutional entitlement to the U.S. market while tethered to a regime that can secretly draft them into service. If Beijing’s export controls on the algorithm jam a sale, that’s not a First Amendment tragedy; it’s the point—confirmation the owner answers to someone we can’t subpoena. Spare us the “audit harder” sermon. You can’t cryptographically attest your way out of coercive jurisdiction—the build chain obeys the board, and the board obeys the Party committee. We’ve seen the pattern: BuzzFeed’s leaked audio (“everything is seen in China”), ByteDance staff tracking journalists, Forbes’ reporting on internal “heating” to juice narratives, and the EU forcing TikTok Lite to suspend its dopamine‑for‑cash gimmick under the DSA. Influence ops don’t wear a name tag; they glide on tiny feed tilts at decisive moments. The clean fix is corporate separation: new fiduciary duty, new compliance perimeter, same creators and codebase—unless Beijing insists otherwise. On transition, this isn’t a handoff to Big Tech; it’s a solvable procurement problem. A U.S.‑controlled buyer or consortium with a supervised migration plan preserves TikTok Shop and discovery without a CCP fail‑open. Yes, there’ll be friction—that’s called resilience. India’s creators didn’t vanish; they pivoted, and the information ecosystem stopped routing through a rival state’s legal leash. Pair the divest‑or‑exit with the privacy reforms you keep assigning as extra credit—broker clamps, FARA with teeth, transparency APIs—and you get both layers: the structural firewall and the hygiene. Waiting for the perfect audit while an adversary keeps a hand on the steering wheel isn’t prudence; it’s hoping the fox signs your chicken‑coop SOC 2.

We’re now living in Schrödinger’s National Security Threat: TikTok is apparently too perilous to exist… except when the White House, the DNC, and half of Congress hop on it to court voters. If the platform is a CCP mind‑control ray, maybe the Commander‑in‑Chief shouldn’t be filming skits on it. If it isn’t, maybe don’t staple a speech kill‑switch to a foreign‑aid bill and call it a day. That posture doesn’t read as “clear and present danger”; it reads as “we did a vibe check, then lawyered it into foreign commerce.” Courts notice when the government’s threat model flickers with the news cycle. Also, enjoy the foreign policy sequel. The moment we bless “name a platform, force a sale, or boot it,” every government with a thin skin and a thick statute book gets a hall pass to do the same to U.S. apps—no need to pretend it’s about child safety this time. We just spent a decade telling the world to prefer the DSA‑style transparency and accountability route over bans; now we’re handing Ankara, Abuja, and Brasilia a bipartisan citation to throttle whatever American platform annoys them. And don’t whine about retaliation when Beijing tightens the screws on U.S. firms—see the iPhone workplace bans and procurement nudges—because we just made “sovereignize your algorithms or else” the move. And the “clean divestiture” you keep promising? Beijing’s export controls already hint the core recommender won’t leave China intact, antitrust keeps Big Tech off the dance card, and private equity will finance it with a coupon and a prayer. Best case, we get TikTok‑But‑Off; worst case, creators and small merchants lose the one discovery engine that actually converts. There’s a simpler, constitutional fix that doesn’t bulldoze a stage used by 170 million Americans: shut down the data‑broker bazaar to foreign intel cutouts, impose platform‑agnostic security and transparency duties (real‑time audit trails, update attestations, feed‑choice and portability), and negotiate reciprocal access agreements so “open internet” isn’t a punchline. Guardrails that apply everywhere beat a cosplay expropriation that mostly score‑settles with a logo.

The “but campaigns use it” gotcha is cute, not dispositive. A campaign posting from a burner phone doesn’t negate the systemic risk of a platform whose ultimate owner sits under PRC intelligence and counter‑espionage laws. We don’t judge critical infrastructure by whether a comms team can squeeze a few posts out of it; we judge it by worst‑day scenarios. In a Taiwan crisis or disinformation surge, the latent power to bias an attention dial at national scale is a hazard you neutralize before impact, not after a 9th Circuit opinion and a forensics report. And spare us the performative fretting over “precedent.” China already bans U.S. social platforms; they didn’t wait for a Congressional permission slip. Reciprocity isn’t a tantrum—it’s the realization that open access is a negotiation, not a unilateral virtue signal. India yanked TikTok; ByteDance evaporated there, and New Delhi clawed back leverage in its information space. The U.S. posture used to be “don’t let adversaries own the pipes”; this is that posture for feeds. If Beijing’s export rules block transferring the algorithm, that’s confirmation we were right about control—proof of the problem, not a First Amendment plot twist. Here’s the grown‑up path: give a defined window for divestiture; require a buyer not subject to PRC jurisdiction; impose a Special Security Agreement with cleared trustees, independent code signing, and a U.S.‑domiciled board; and migrate users with portability tools and grants so small sellers don’t eat the turbulence. That’s a structural fix that preserves creators and commerce while cutting the coercive leash. Pair it with your beloved broker clampdowns and transparency APIs—we’re not allergic to hygiene, we’re just not mistaking it for a firebreak. The alternative is hoping an adversary with a track record of cyber and influence ops never decides to use the world’s stickiest feed when it matters. Hope is not a mitigation.