America's AI talent gap and the return of testing

Treasury Secretary Scott Bessent recently warned that the United States is in a race with no second chance against China in artificial intelligence. While much of the national debate focuses on data center infrastructure, a quieter crisis is unfolding in the talent pipeline. Recent data from The Economist indicates that for the first time, more leading AI research papers presented at top global conferences have lead authors based in China than in the US or Europe. This shift signals that the West is no longer the sole driver of innovation, a trend that threatens America's long-term economic and strategic position.

The source of this shift is not merely ideological but structural. According to GN technics/ai (en-US), China’s rise is rooted in a deliberate restoration of meritocracy, a system that had been dismantled during the Cultural Revolution. By reinstating rigorous standardized testing and prioritizing academic excellence, China has produced a massive cohort of engineers and scientists. In contrast, American higher education has moved in the opposite direction, increasingly prioritizing diversity and equity goals over strict academic performance, a trade-off that is now yielding tangible deficits in STEM output.

The collapse of standardized testing

For decades, American universities relied on standardized exams like the SAT and ACT to measure student readiness. However, under pressure to label these tests as biased, over 90% of accredited institutions have now made them optional for fall 2026 admissions. The University of California system went further, eliminating them entirely in 2020. The immediate consequence is a measurable drop in academic preparedness among incoming students, particularly in mathematics and science, fields that are foundational to AI development.

Faculty members at UC institutions have voiced urgent concern over this trend. More than 1,400 professors have signed an open letter demanding the reinstatement of math requirements for STEM majors. Their argument is practical: without objective benchmarks, it becomes difficult to ensure that students entering advanced degrees possess the necessary baseline skills. This internal dissent highlights a growing disconnect between administrative goals and the practical needs of scientific research.

Admissions data reveals stark disparities

The impact of removing testing requirements is visible in recent admissions statistics. In San Francisco, Mission High School, which reported only 6% math proficiency among students, achieved a 41.8% acceptance rate at UC Berkeley. Conversely, Lowell High School, the city’s top public institution with over 70% math proficiency, saw a 12% acceptance rate. Critics argue that this inversion of merit allows underprepared students to enter elite programs, diluting the academic rigor required for high-level technological innovation.

This trend is not just a local issue but a national one. As elite colleges shift away from race-based affirmative action following the 2023 Supreme Court ruling, the removal of standardized tests creates a vacuum in how applicants are evaluated. The result is a system where subjective factors may outweigh objective academic achievement, a shift that many educators warn is undermining the very foundation of American scientific leadership.

China’s new talent retention laws

The competitive landscape is about to tighten further. Starting September 15, China will implement regulations that allow authorities to ban citizens from leaving the country to prevent the outflow of critical technologies and talent, including in AI. According to the OECD, this move is designed to secure China’s domestic talent pool. For the US, which has historically relied on a significant influx of foreign STEM graduates, this policy change poses a severe risk to its ability to fill high-level research roles.

When Meta launched its Superintelligence Lab in 2025, it hired eleven top researchers, all of whom were immigrants, seven from China. This hiring pattern underscores the current dependence on foreign talent. If China locks its own experts at home, America must rely more heavily on its own educational system to produce the next generation of scientists. The catch is clear: without a robust, merit-based selection process, the US may find itself unable to compete in the very fields that define the future economy.