How an 'Ethnic Unity Law' Reaches Into Taiwan: Beijing Moves From Economic Pressure to Cross-Border Lawfare

The place this dispute is most easily distorted is where three different levels of statement get kneaded into one sentence.

The first layer is the text of the law itself. China's National People's Congress passed the Law on Promoting Ethnic Unity and Progress on March 12, 2026; it took effect on July 1. The law as a whole writes "forging a strong sense of community for the Chinese nation" into education, media, the internet, enterprises, social organizations, risk monitoring, legal liability, and Taiwan work. ^[1]

The second layer is official interpretation. Beijing describes it as the foundational law of ethnic affairs work, one that converts the Chinese Communist Party's theory of ethnicity into the will of the state, and denies that extraterritorial accountability amounts to long-arm jurisdiction. ^[4][15] Taiwan's Executive Yuan and Mainland Affairs Council, by contrast, take the view that it pushes cross-strait policy from "opposing independence" toward "promoting unification," folds Taiwan into ethnic affairs work and cross-border accountability, and could pressure people engaged in exchanges into self-censorship. ^[2][6][7]

Only the third layer is reasonable inference: how might these provisions, combined with the 2024 opinion on punishing independence, China's Criminal Law, and national-security and internet regulations, affect Taiwanese people and organizations? Inference can be made, but it must come with conditions attached — "could" must not be written as "already has."

For example, the Executive Yuan says the new law places an obligation on the Taiwanese people to uphold China's unification — this is the Taiwan government's overall assessment, not the literal wording of Article 10. What Article 10 actually says is that "citizens of the People's Republic of China" have an obligation to safeguard national unification, sovereignty, security, and development interests. ^[1] The fact that Beijing does not recognize Taiwan's statehood does mean Taiwan cannot rely on this wording alone for reassurance; but a rigorous piece of writing still cannot rewrite the statute to say it explicitly orders "all Taiwanese people" to support unification.

Likewise, the law does not explicitly classify Taiwanese people as an "ethnic minority." It uses the political-identity language of "Taiwan compatriots" and "the Chinese nation." This distinction is not mere pedantry over wording — it is the starting point for judging how the law actually operates.

The provision most directly connected to Taiwan in the entire law is Article 21.

It requires Chinese state organs to promote cross-strait economic and cultural exchange and cooperation, advance integrated development, and strengthen Taiwan compatriots' "sense of belonging, identification, and pride" in the Chinese nation, along with the understanding that "compatriots on both sides of the strait are Chinese and belong to the same Chinese nation." ^[1]

The text does not say that a Taiwanese person who disagrees will be sentenced to a certain number of years, nor does this article establish a crime called "failing to identify with the Chinese nation." Writing it up as straightforward criminal law would exaggerate the statute.

But treating it as a toothless declaration about exchange is too light a reading as well.

Because Article 21 does not stand alone. Article 16 runs the sense of community through education; Article 19 requires news media, publishing, and internet services to bear propaganda responsibilities; Article 20 pushes the requirement into family education and work with minors. By Article 44, mass organizations such as the All-China Federation of Taiwan Compatriots are likewise required to carry out community-building work. ^[1]

What it changes is the legal purpose of exchange. Cross-strait cultural, educational, religious, youth, and commercial exchanges are, under this law, no longer simply person-to-person contact; they are also assigned the national task of shaping the identity that "both sides of the strait are Chinese." For Beijing, this is part of national rejuvenation and reunification; for a Taiwan that does not accept the People's Republic of China's jurisdiction, it is stuffing political loyalty into the order of exchange.

This does not mean every exchange will be forced to swear an oath. The more reliable judgment is that Chinese authorities, regulated institutions, and partner organizations will now have a firmer legal basis for demanding that event content, organizational charters, or messaging conform to the community narrative. The pressure need not appear first as a criminal case — it is more likely to appear first as cancelled invitations, edited-out content, conditions attached to cooperation, and self-censorship. This is a reasonable inference, not a case that has already happened.

If Article 21 deals with Taiwanese identity, Article 45 deals with organizations.

It splits into two categories of requirement: enterprises and public institutions should incorporate the forging of a strong sense of community for the Chinese nation into business training, cultural-building activities, and the like; social organizations such as trade associations, chambers of commerce, learned societies, and foundations should embody the requirement in industry self-discipline codes, professional ethics standards, or organizational charters. ^[1]

Here too, two extreme readings must be avoided.

One claim is that, from July 1 onward, every Taiwanese company must write "the community of the Chinese nation" into its employee handbook. This is not correct. A Chinese law does not gain universal enforceability inside Taiwan's territory simply by being published in an NPC bulletin.

The other claim is that Taiwanese companies need not pay any attention at all, because this is purely a matter of Chinese domestic law. This also ignores the reality of cross-border operations: Taiwanese businesses with subsidiaries, factories, offices, or platform accounts in China, and Taiwanese organizations partnering with Chinese schools, associations, or religious or cultural institutions, could all receive this requirement through their local legal entity, licenses, employee training, website content, or cooperation contracts.

Legal risk here is not determined by passport alone — it is determined by the contact points. The more assets a company has in China, the more its data depends on Chinese platforms, the more frequently its key executives travel there, and the more its partners are subject to official regulation, the greater the leverage behind any demand for compliance.

Article 31 then draws internet platforms into the same net. Where a service provider encounters content determined to undermine ethnic unity, it must stop transmitting it, delete it, retain records, and report to the competent authorities. ^[1] This means an article that is perfectly legal in Taiwan, once it moves through a Chinese platform, is sent to Chinese employees, or enters a Chinese company's internal network, no longer faces simple market preference — it faces a compliance judgment made by a platform or an employer.

The phrase "from economic pressure to cross-border lawfare" in the headline can easily create the impression that Beijing has switched weapons: tariffs before, law now.

The reality is not replacement — it is accumulation.

In 2024, China suspended, in batches, tariff concessions on some ECFA products. China's Ministry of Commerce described the measure as retaliation after Taiwan violated ECFA and adopted discriminatory trade restrictions; Taiwan's Mainland Affairs Council said that China's expansion to a further 134 products that June amounted to economic coercion. ^[8][9] During the WTO's trade policy review of China that same year, multiple members also expressed concern about economic coercion and the use of economic measures for political purposes. ^[10]

On June 21, China's Supreme People's Court, Supreme People's Procuratorate, Ministry of Public Security, Ministry of State Security, and Ministry of Justice jointly issued a 22-point opinion on punishing "diehard Taiwan independence separatists." That document lists, more directly than today's new law, which political conduct toward Taiwan can be treated as splitting the country, and discusses trials in absentia, wanted notices, disposal of property, and severe criminal penalties. ^[5]

This timeline matters because it forecloses two mistakes.

First, cross-border lawfare did not suddenly begin in July 2026. Beijing has long used criminal statutes, watchlists, wanted-notice language, and national-security mechanisms to pressure Taiwan.

Second, the ethnic unity law is not a repeat of an old document either. The anti-independence opinion focuses on political conduct that Beijing defines as splitting the country; the new law places education, culture, religion, media, enterprises, social organizations, and individuals abroad all into a broader "ethnic unity" and community framework.

Economic tools directly change costs; criminal and national-security tools intimidate specific targets; the ethnic unity law redefines the environment of compliance and loyalty. All three can operate simultaneously: market access and licensing requirements first compel an organization's cooperation, platform rules then remove content, and criminal or national-security norms pursue accountability where necessary. Law does not replace economics — it supplies economic pressure with more political conditions.

Article 63 is the single sentence at the center of the controversy: organizations and individuals outside China's territory that engage in conduct such as undermining ethnic unity or creating ethnic division against China shall be held legally liable in accordance with law. ^[1]

The text of the law does explicitly bring in organizations and individuals outside China's borders, with no nationality restriction. Saying it has no extraterritorial orientation at all does not match the wording.

But "China claims it can hold someone liable" and "China can enforce this anywhere in the world" remain two different things.

Article 63 does not, in the same provision, set out an independent offense, its elements, a penalty range, an extraterritorial evidence-gathering procedure, or an extradition mechanism. Article 58 says only that a person who undermines ethnic unity shall be ordered by the relevant department to make corrections and shall be handled under other laws and administrative regulations; where the conduct constitutes a violation of public-security administration or a crime, it shall then be punished according to law. ^[1] The new law functions more like an interface — actual sanctions still have to connect back to the Criminal Law, public-security regulations, national-security law, internet regulations, or other norms.

Hu Weilie, a vice minister of China's Ministry of Justice, denied at a State Council Information Office press conference that this amounts to "long-arm jurisdiction," calling it a legitimate act to safeguard sovereignty, security, and development interests and consistent with international practice. ^[15] This is Beijing's official defense. Taiwan's government, for its part, calls it long-arm jurisdiction and cross-border repression. ^[2][7] This is Taipei's official risk assessment.

A neutral legal judgment should take one further step: states routinely assert criminal jurisdiction over harms committed abroad, but whether such a jurisdictional claim can actually be enforced depends on where the person is, where the assets and data sit, judicial-assistance arrangements, and whether other countries recognize the claim. China cannot use Article 63 alone to order Taiwan's police to make an arrest in Taipei, nor can it require every foreign court to automatically recognize its judgments.

But if the person enters China, Hong Kong, or Macau; if a company's assets and staff are in China; if the data sits in the hands of a Chinese platform or partner — the obstacles to enforcement fall away. The real force of cross-border lawfare usually lies not in a foreign government arresting someone on Beijing's behalf, but in whether Beijing can control travel, assets, data, and market access.

Article 63 is therefore not a global arrest power, but neither is it a meaningless paper tiger. What it adds is a political and legal interface that lets Beijing describe speech, organizations, and activities outside its borders as harms to China's ethnic unity, and then deploy other tools depending on the available contact points.

This point has to be stated on its own.

As of this article's press deadline, there is no publicly verifiable case in which a Taiwanese individual, company, or group has been detained, prosecuted, sentenced, wanted, or had assets frozen with the Law on Promoting Ethnic Unity and Progress explicitly cited as the core basis. The new law only took effect on July 1; extrapolating a future rate of enforcement from less than two weeks of blank record is not sufficiently evidentiary.

China's past handling of Taiwanese individuals under the crime of splitting the state, or its naming of "diehard Taiwan independence separatists" under the 2024 22-point opinion, can prove that Beijing already has legal tools aimed at Taiwan — but neither can be passed off as a case brought under the new law. ^[5]

No case does not mean no risk. A law's chilling effect can well appear before any judgment does; a company cancelling an invitation, a platform deleting a post, a researcher changing a topic, or a participant not daring to travel to China may never leave behind a court docket number. But conversely, risk cannot be filled in by imagination either. An article that claims a particular Taiwanese traveler has already been detained under the new law over the contents of their phone, yet cannot produce a name, a date, and an official document, is fiction.

Responsible monitoring should wait for several verifiable signals: the first time a Chinese law-enforcement or judicial body cites this law in an official document; the first time the Taiwan Affairs Office or another department names a specific Taiwanese target under this law; the first time a platform, company, or group invokes Article 45 or Article 31 to demand a Taiwanese partner change its charter or delete content; and how a court uses Article 58 to connect it to other laws.

Until those signals appear, the most accurate sentence is this: the legal risk has been declared; the actual pattern of enforcement still awaits evidence.

The side making the strongest defense of Beijing's argument would say that Taiwan's government has repackaged a domestic ethnic-governance law as a global manhunt order. Article 10 names only PRC citizens, Article 21 carries no criminal penalty, and Article 63 cannot pierce foreign sovereignty either. Every country's national-security law has some form of extraterritorial application; if the mere appearance of the words "outside the territory" is enough to call something long-arm jurisdiction, legal analysis degenerates into a partisan label.

This side would also point out that the main criticisms of the new law from UN special procedures, Amnesty International, and legal analysis focus on the assimilation of Uyghurs', Tibetans', and Mongolians' language, religion, education, and cultural autonomy. ^[11][12][13] That material can reveal a governance logic, but it cannot be directly treated as a case of Taiwanese victimization. If Taiwan uses risk as a pretext to block exchange across the board, it too could expand government control and turn every ordinary interaction into a security matter.

These points stand.

But the other red team would ask back: if this were purely a domestic law, why does Article 21 specifically write in "Taiwan compatriots," and Article 63 specifically write in organizations and individuals outside the territory? If the absence of an independent penalty range means it has no force, why does Article 58 need to connect liability back to other laws? China already has its anti-independence opinion and national-security criminal law — a framework law's function was never going to require creating every single penalty from scratch.

This side cares more about the "first-case problem." If Taiwan waits until the first person is detained or the first company is forced to hand over data before acknowledging that the provisions carry risk, the cost is borne by the individual. Legal uncertainty by itself can change behavior, especially when China leaves the political line to be interpreted by law-enforcement and Party-state organs.

These points stand as well.

A neutral synthesis is not each side splitting the difference — it is answering two separate questions. Can China enforce this directly inside Taiwan's territory? Normally not, unless Taiwan's own institutions cooperate. Can China pressure a Taiwanese target that has contact points involving its personnel, assets, data, or platforms? Yes — and the more contact points there are, the stronger the leverage.

UN special procedures sent a letter to China in April raising concerns, from the angles of minority rights, cultural rights, education, freedom of expression, assembly, and religion, that the new law could institutionalize assimilation measures. Analysis from Amnesty International and the Council on Foreign Relations likewise points to trends toward a common national language, community-oriented education, and the weakening of ethnic autonomy. ^[11][12][13]

These criticisms concern, first and foremost, Tibetans, Uyghurs, Mongolians, and other groups within China who have long lived under high-pressure governance. Treating their experience merely as a metaphor for Taiwan's risk is disrespectful to the victims and would also confuse the legal facts.

But the governance logic has comparative value. The new law does not only prohibit violent separatism — it treats language, education, family, media, religion, culture, and enterprise all as sites of community-building. When that same logic is carried into Taiwan policy through Article 21, what Taiwan faces is no longer only a territorial claim to unification, but the question of "which identity, which language, which historical narrative counts as legitimate exchange."

A European Parliament delegation also expressed concern about this law when resuming interparliamentary exchanges with China. ^[14] The point of internationalizing this issue should not be to demand that foreign countries accept every adjective in the Taiwan government's account wholesale, but to lay out verifiable questions clearly: does the text satisfy requirements of clarity and proportionality? Do parties outside the territory have any remedy? What are platforms and companies being asked to do? What kind of speech has been deleted or turned into a case?

Only if Taiwan can make its evidence more granular than its slogans does the issue of cross-border repression stand a chance of moving from a cross-strait political quarrel into international human-rights and rule-of-law cooperation.

The Executive Yuan has put forward five categories of response: raising literacy and civil-servant training, strengthening review of cross-strait exchanges, improving prevention and countermeasures, providing assistance to victims, and cooperating with like-minded countries. ^[2] The overall direction is comprehensive; in execution, it still needs to avoid turning risk management into indiscriminate blockage.

For the state, the first task is not to issue yet another press release that says only "stay alert," but to publish a provision-by-provision legal matrix — which parts are statutory text, which are Beijing's interpretation, which are Taiwan's inference, and which already have cases — and keep it continuously updated. Travel advisories should also be tiered: ordinary tourism, academic research, media reporting, religious activity, political advocacy, and people already named by Beijing should not all receive the same single warning. If detention occurs, or a device is demanded, or a political statement is required, who families, companies, and government agencies should contact must be made available in advance.

For industry and social intermediaries, associations, schools, and religious and cultural groups need to inventory four contact points: whether personnel travel to China, Hong Kong, or Macau; whether assets and legal entities are located in China; who controls the data; and which platform their activities and content depend on. What they need are contract clauses, incident-reporting procedures, and legal support — not a registry of members' political loyalty.

For SMEs, the most practical step is not hiring an expensive geopolitical consulting team, but a one-page red-flag checklist: does a device traveling to China store unnecessary personnel data? Does the local partner require signing on to a political commitment? Are key accounts and backups all placed on a single Chinese platform? If an employee is detained or summoned for questioning on short notice, who is the 24-hour point of contact? Can a subsidiary in China keep operating without implicating the Taiwanese parent company's data?

These measures are not put in place because every Taiwanese businessperson will be caught — they exist because low-probability, high-damage events need contingency plans. Nor is this asking companies to self-censor; it is letting them know when they should refuse, isolate data, adjust travel, or withdraw from cooperation.

Economic pressure deals with price. When Beijing withdraws tariff concessions, a company can calculate the cost, switch markets, and adjust its supply chain. Traditional criminal intimidation deals with the person named: whoever is designated a "diehard Taiwan independence separatist" bears risks to travel, assets, and reputation.

The ethnic unity law takes a step further. It does not only ask what you have done — it also tries to prescribe which identity exchange, education, enterprises, and organizations should be cultivating. Article 21 says Taiwan compatriots should strengthen their sense of belonging, identification, and pride in the Chinese nation; Article 45 requires organizations to incorporate the sense of community into their institutions; Article 63 keeps open an entry point for pursuing conduct outside the territory. ^[1]

This is exactly where cross-border lawfare is most worth being vigilant about — and also most easily exaggerated.

It is worth being vigilant about because the boundaries of identity and thought are blurrier than a tariff schedule, and organizations, for the sake of market access and personnel safety, may self-censor even before any court order arrives.

It is easily exaggerated because Beijing's legal declaration does not automatically pierce Taiwan's sovereignty, and because we still do not have a case applying the new law to Taiwan that could prove at what speed, frequency, and severity it will actually land.

A mature Taiwanese response cannot choose between "nothing to worry about at all" and "a global manhunt." It has to place the risk back onto observable contact points: who enters a space China can control, what assets remain there, which data passes through which platforms, and which partner might be asked to comply. It then reduces vulnerability point by point, through public legal analysis, tiered advisories, corporate contingency plans, case-by-case assistance, and international evidence cooperation.

What Beijing really wants to expand is not only the jurisdictional radius of one law, but the political radius of "who has the authority to define who the Taiwanese people are."

Taiwan does not need to exaggerate that power on Beijing's behalf, nor can it pretend to itself that this radius does not exist.