America’s Foreign-Student Crisis Is Thailand’s Opportunity

America’s foreign-student machine is stalling at exactly the moment Thailand has an opportunity to move up the education value chain. The United States still reports more than a million international students, but that reassuring headline is a rear-view-mirror statistic. The people closest to the top of the funnel—companies that recruit and place foreign students for American universities—describe the current market in less academic terms. They call it a bloodbath.

I work closely with several of these companies. They report a collapse in inquiries, applications and admissions, driven by four concerns that now reinforce one another: uncertainty over whether a student visa will be issued, uncertainty over whether a graduate will be allowed to work afterward, a political climate perceived as hostile, and the cost of spending four years in the United States. These are not yet public national statistics. Universities do not report admissions through a common real-time system, agents treat conversion data as commercially sensitive, and the official series arrive too slowly to capture a market turning between recruitment seasons. The absence of public data should be stated plainly. It should not be mistaken for evidence that nothing is happening.

The numbers we do have point in the same direction. Open Doors reported a record 1,177,766 international students in 2024–25, up 5 percent, but new international enrollment fell 7 percent to 277,118. The preliminary fall 2025 survey was worse: new enrollment fell 17 percent and graduate enrollment 12 percent, while undergraduate enrollment rose 2 percent and Optional Practical Training rose 14 percent. F-1 visa issuance from May through August 2025 was reported 36 percent below the same months a year earlier. Among institutions reporting a decline in new students, 96 percent cited visa-application concerns.  That is as close as the public data come to the market intelligence. It is close enough.

The apparent contradiction between record enrollment and collapsing interest is mostly an artifact of stocks and flows. A four-year undergraduate degree contains four entering classes. In the first year of an admissions shock, roughly three quarters of the students were already in the system; they continue moving toward graduation even if the incoming class falls sharply. A 40 percent decline in first-year intake would initially reduce the four-year undergraduate stock by only about 10 percent, all else equal. Add graduate programs of different lengths and nearly 300,000 people on OPT—students counted in the Open Doors total even though they are no longer sitting in classrooms—and the headline number becomes heavily insulated from current demand. Total enrollment is not false. It is simply the wrong instrument for detecting a sudden break in admissions.

This matters because international education is an export industry. Foreign students bring money earned elsewhere into American tuition accounts, apartment buildings, grocery stores, insurers, airlines and university towns. NAFSA’s model puts their 2024–25 economic contribution at $42.9 billion and 355,736 supported jobs. The national figure is large, but the institutional exposure is concentrated. International students often pay nonresident or near-sticker tuition, particularly in professional master’s programs in computing, engineering and business. The most vulnerable schools are not Harvard or Stanford. They are regional public universities and tuition-dependent private colleges that built budgets around a steady flow of full-paying foreign graduate students. The University of Central Missouri, where international students had been about thirty percent of enrollment and nearly a quarter of tuition revenue, already reported a 5.8 million dollar drop in net tuition tied to lower graduate enrollment. The University of North Texas watched international graduate enrollment fall from about 6,200 to just under 3,400 in a single year. These institutions feel an admissions collapse first, because their students move through fastest.
The obvious question is where the next generation of internationally minded Asian families goes instead. Thailand’s current strength sits one stage earlier in the chain—and that is precisely why the opportunity is interesting.

Thailand’s Missing University Layer
Thailand has quietly become one of Asia’s strongest platforms for international K–12 education. The Thai Ministry of Education counted 81,057 students in international private schools (K-12 equivalent) in 2024, including 45,155 in Bangkok. A separate official-sector count found 249 international schools and more than 77,700 students; the definitions differ, but the scale is clear. Bangkok has the premium schools, counsellors and expatriate infrastructure. Chiang Mai has become a relocation center for Chinese families seeking English-medium education and relief from the pressure of China’s domestic system. Phuket combines boarding schools, tourism infrastructure and lifestyle. Chonburi and Sriracha serve Japanese industrial families whose parents work in Thailand’s manufacturing economy.
This is more than cheap tuition. Thailand offers a complete family product: international curricula, English immersion, housing, private healthcare, guardianship, visas, domestic help, transport, test preparation and a social environment in which an East Asian family can move abroad without moving to the other side of the world. It also offers an unusually broad set of educational identities. A family can choose British, American, International Baccalaureate, Chinese, Japanese or hybrid schooling inside the same national market.

The schools are good at what families hire them to do. NIST International School has more than 1,800 students from dozens of nationalities; its class of 2025 received 296 offers from 188 universities in 19 countries, including Oxford, Cornell, the University of Pennsylvania, Melbourne, McGill and Toronto. Bangkok Patana runs the British curriculum into the IB and a global university-counselling pathway. Prem and UWC Thailand combine boarding, English support and IB preparation with overseas progression. Thailand has built an effective launchpad into the Western university system.

That success also reveals the leak. The product is designed to prepare students in Thailand and monetize their university years somewhere else.

Thailand had 34,202 international higher-education students in the first semester of 2022, including 21,419 from China. That is a meaningful market, but it is much thinner and less internationally diversified than the school ecosystem. The K–12 and university counts cover different years and definitions and should not be turned into a retention rate. No reliable national retention rate exists. Still, the architecture is visible in the schools’ own destination lists: Thailand attracts families, educates their children, gives them English and internationally recognized qualifications, then celebrates when the most valuable four years of the relationship are exported to the United States, Britain, Australia or Canada.

For decades that was not a failure. A Western university place was the final product. The Thai school sold the preparation, and the family willingly paid the destination country for the credential, network and work option. What has changed is the stability of that destination offer. If the United States cannot assure a student that a visa will arrive on time, that the political environment will remain tolerable, or that a pathway to employment will exist after graduation, the family must reconsider how much of the degree needs to be purchased there.

Thailand does not need to replace the American university. It needs to retain one or two more years of the student.

The most plausible model is a modular degree: one or two years in Thailand, followed by one or two years at an affiliated foreign university; or a foreign degree delivered largely in Thailand with a semester abroad; or a Thai international-college degree paired with foreign accreditation, exchange and multinational internships. The family buys the Western brand and progression option without paying Western living costs for four years. The foreign university gains a prepared cohort without carrying the full recruitment burden. Thailand keeps tuition, housing, healthcare and family spending that currently leave the country.
Pieces of this model already exist. Mahidol University International College’s Preparation Center has offered English and mathematics preparation with conditional progression into MUIC since 1998. Thai-Nichi Institute of Technology connects Japanese-language and technical education to exchange and company partners. The Asian Institute of Technology has offered Japan-linked scholarships with employment attached. What Thailand lacks is not institutions; it lacks a sufficiently broad, legible system of credit transfer and affiliated degrees tied directly to the international-school pipeline.

Vietnam shows what the next layer can look like. Ton Duc Thang University advertises 2+2 (2 years in country/2 years out), 3+1 and related joint-training structures with foreign partners. National Economics University offers 4+0, 3+1 and 2+1 routes with institutions including Andrews University and the University of the West of England. Fulbright University Vietnam has developed exchange and 3+2 (with a Master’s) pathways with American partners. RMIT Vietnam demonstrates the full foreign-campus model, although that is the expensive end of the spectrum. These programs are in various stages of implementation, with some still just MOUs, but the direction is present. Vietnam has the larger outbound-student pool and a lower cost base; Thailand has the stronger international-school pipeline, family-relocation ecosystem and lifestyle brand.

The winning program would not be another generic “international” business degree with a foreign logo on the brochure. It would publish the awarding institution, accreditation, credit map, transfer conditions and refund terms. It would offer several exits rather than one brittle promise: complete the degree in Thailand, transfer to a partner abroad, or combine a local degree with a foreign master’s. It would concentrate initially on fields that travel well across borders—business, computing, engineering, hospitality and non-licensure health programs—and pair them with internships at the Japanese, Chinese, Korean, European and American companies already operating in the region.

Then there is artificial intelligence, which turns this from a cost argument into a question of credibility.
Universities have spent centuries treating the written product as evidence of the mind that produced it. Much of American higher education is deeply attached to essays, take-home papers, projects and continuous assessment, though the mix varies enormously by institution and discipline. Generative AI has broken the link between a polished submission and the student’s independent capability. The latest surveys are stale almost as soon as they appear: a 2025–26 British survey found 95 percent of undergraduates using AI in some form, while a large American study using 2023–24 data found two thirds had used it and estimated AI-assisted cheating among users at 9 percent. Professors’ lived experience is often more alarming than either number. They report classes awash in machine-written work and, in some cases, sharp grade declines when assessment is moved back into the room.

Those faculty reports are not yet a national dataset, and the distinction matters. But the economic problem does not depend on proving that cheating is 9 percent, 30 percent or 60 percent. If an employer or graduate school cannot know whether the student wrote the paper, the paper has already lost value as a signal. AI detectors do not solve the problem: they can be evaded, generate false positives and may perform differently on non-native English writing. The only durable answer is to verify capability directly.
American universities will adjust. They will bring back more supervised examinations, in-class writing, live coding, practical demonstrations and oral defenses. Australia’s universities have been quick to respond: the University of Sydney’s two-lane model—secure in-person assessment for unaided capability, open assessment when AI is permitted—is a sensible template, as is James Cook University’s use of short vivas, code walk-throughs and process logs. But redesigning thousands of courses takes time. Large lectures need rooms and invigilators. Oral examinations need faculty hours and accessibility rules. Departments must decide what students are still expected to do without AI and what they are expected to do with it. Governance moves by committee; technology moves by semester.

During that interval, the conventional degree is not worthless, but part of its signaling value is under review. The vulnerability is greatest at expensive, mid-ranked universities whose offer consists largely of lectures, take-home assignments and an uncertain work visa. Elite brands, laboratories, clinical training, research networks and regulated credentials remain scarce. Generic academic content does not.
Many Asian universities begin this transition closer to the destination. It would be too simple to say that America uses papers and Asia uses exams: universities in China, Japan, Korea, Singapore, Malaysia, Thailand and Vietnam all use mixtures of examinations, coursework, reports, projects and presentations. But supervised final examinations, attendance thresholds and proctored testing remain more visible in the published rules of many Asian institutions. Peking University and HUST explicitly provide for written, oral and practical assessment; Japanese and Korean universities retain structured in-class examination systems; Singaporean and Malaysian universities commonly combine continuous assessment with examinable components. They face the same AI problem, but many have less institutional distance to travel before a degree once again certifies observed performance.

That creates a second opening for Thai and Vietnamese hybrid programs. A foreign-affiliated degree delivered in Southeast Asia can be cheaper than four years abroad and, if designed properly, more credible than a paper-heavy course transplanted online. The regional campus can provide supervised exams, vivas, laboratories, group work observed by instructors and internships with local multinationals. The foreign partner supplies the curriculum, quality assurance, brand and progression route. AI can be used openly for tutoring, translation and research assistance, while the institution verifies what the student can do without it and what the student can responsibly do with it. The problem here is implementation, and a desire for US and foreign-affiliated programs to be identical, bolstering their legitimacy at the cost of innovation.

This is not an argument for recreating the Asian cram school at university level. Closed-book exams can be secure and still measure the wrong thing. The stronger model combines observed performance with practical work: an exam to verify knowledge, a viva to verify understanding, a project to verify application and an internship to verify that the student can function inside an organization. In the age of AI, the premium product is no longer content. It is trustworthy evidence of capability.

Thailand is unusually well placed to sell that product because it already has the customers. The families are in Bangkok, Chiang Mai, Phuket and Sriracha. The schools, counsellors, hospitals, apartments and relocation companies already serve them. The students already study in English and prepare for foreign universities. The missing piece is a university layer strong enough to retain them without closing off the world they came to access.

America’s foreign-student shock does not automatically become Thailand’s gain. Markets do not transfer themselves. Thailand must build the partnerships, credit systems, assessment standards and employer links that turn a school destination into an education economy. But the opportunity is now visible. Thailand has already built the first half of the international pathway. The next business is to stop exporting the second.