The United States has historically led the world in technological innovation through its internationally renowned education institutions, innovative industries, top-tier research laboratories and, critically, its unique ability to attract talent worldwide. 

Immigrants play a key role in sharpening America’s technological edge.¹ In recent years, the demand for artificial intelligence talent has greatly exceeded domestic supply, leading to a large share of foreign-born AI students, workers and entrepreneurs in the United States.² Although important, efforts to increase the domestic AI workforce are insufficient to fill the immediate demand for AI talent. At the same time, other countries are developing their own capabilities and institutions in AI and aggressively recruiting AI talent through new immigration policies.³ In this competitive environment, current U.S. immigration policies, many of which date back decades, may work against the country’s historic strength in attracting and retaining international talent. 

Although various factors shape any country’s AI competitiveness, this paper will focus solely on immigration policies relevant to AI talent in the United States and four sample economic competitor nations: the United Kingdom, Canada, France, and Australia. These countries were selected for their unique policies on attracting international talent in AI and other tech fields. The United States may be able to evaluate lessons learned from these countries to inform domestic policy. On the other hand, if the United States fails to adapt to an increasingly competitive global technology talent landscape, other countries may begin to draw AI talent away from American schools and employers. 

This paper analyzes policies relevant to four categories of immigrants: students in AI-related fields of study, workers in AI-related industries, distinguished AI workers (that is, individuals internationally renowned for their achievements in AI), and AI entrepreneurs. These groups represent the range of backgrounds and experience levels that nations need to compete in AI. We explore trends that, taken together, may be making the U.S. immigration system less attractive to these groups relative to other countries’ systems: 

• Within the last five years, the UK, Canada, France, and Australia have adopted major immigration reforms to attract talent in AI and other technical fields. The United States has not.
• Despite growing job opportunities, recent graduates and others may be restrained from contributing to the U.S. AI workforce to their full potential—partly due to current caps, backlogs, and sponsorship processes at the expense of the employer for temporary work visas and permanent residency. In contrast, Canada’s new immigration policies quickly bring in skilled migrants and integrate graduates into the workforce. The UK is proposing similar changes to ease and expedite the immigration process for technically skilled migrants.
• The United States’ per-country quotas on permanent residency—which remain unchanged for decades—have created a significant bottleneck, especially for Indian nationals who make up the 25 percent of Silicon Valley’s technical workforce. The other countries in this analysis forego quotas on permanent residency status and allow immigrants who meet permanent residency requirements to apply. 
• Although data is scarce and important trends are in nascent stages, empirical indicators suggest that some new AI-focused immigration policies in other countries are successful.
• The United States has long attracted immigrant entrepreneurs with its innovative culture, but does not offer an entrepreneur visa. Entrepreneur visas offered by other countries in this analysis have largely failed due to unrealistic and vague metrics for business success or long processing times. The United States could learn from the mistakes of competitor countries and design its own visa to increase its immigrant entrepreneur population and create jobs for Americans. 

The United States historically and presently benefits from a strong baseline of technological innovation through existing institutions. Immigration trends alone are unlikely to eliminate the U.S. advantage in the near term. However, the global landscape is shifting, and restrictive immigration policies threaten to undermine U.S. AI progress in the long term. To ensure America remains competitive for AI talent in the coming years, U.S. policymakers should consider reforms to current immigration statutes, regulations, and agency guidance. Other countries’ immigration reforms suggest three main lessons: 

• Improve temporary visa options for skilled workers. The structure of the H-1B temporary work visa prevents AI talent from contributing to the United States to their full potential. In particular, workers seeking permanent residency while on H-1B status—a process that could take years or decades—would need another sponsorship if they wished to switch positions or employers. Further, the H-1B lottery occurs only once a year, forcing employers to wait until the annual draw on April 1 to learn whether critical employees have been selected. In contrast, Canada has no cap on the number of work permits that can be provided year-round and issues these permits in as little as two weeks. 
• Expand opportunities for permanent residency. Allocating permanent residency status based on decades-old caps and immigrants’ countries of origin, rather than the skills they bring to the United States, has created a bottleneck for highly skilled AI workers who wish to contribute to the U.S. AI workforce in the long term. Specifically, the employment-based green card wait time for Indian nationals, who make up 25 percent of Silicon Valley’s technical work force, is 89 years. There are no formal caps or quotas on permanent residency in the UK, France, Australia, and Canada. Instead, immigrants are eligible to apply once they have lived or worked in the country for a set number of years. 
• Expand opportunities for entrepreneurs. While each of the other four countries in this analysis offers some form of an entrepreneur visa, initial results suggest they have been relatively unsuccessful. The United States should strengthen AI innovation by adopting an entrepreneur visa informed by the flaws of competitor nations to better attract and retain AI entrepreneurs. 

1. We use the term “immigrants” colloquially in this paper, referring to individuals who originate outside the United States and are studying or working in the United States. U.S. immigration law distinguishes between permanent “immigrants” (i.e., green card holders) and temporary “nonimmigrants.” See “What is the Difference Between and Immigrant Visa vs. Nonimmigrant Visa?,” U.S. Customs and Border Protection, https://help.cbp.gov/app/answers/ detail/a_id/72/~/what-is-the-difference-between-an-immigrant-visa-vs.-nonimmigrant-visa-%3F.
2. Zachary Arnold, Roxanne Heston, Remco Zwetsloot, and Tina Huang, “Immigration Policy and the U.S. AI Sector,” (Center for Security and Emerging Technology, September 2019), https://cset. georgetown.edu/wp-content/uploads/CSET-Immigration-Policy-and-the-U.S.-AI-Sector-1.pdf.
3. Daniel Castro, Michael McLaughlin, and Eline Chivot, “Who is Winning the AI Race: China, the EU, or the United States?” (Center for Data Innovation, August 19, 2019), https://www.datainnovation. org/2019/08/who-is-winning-the-ai-race-china-the-eu-or-the-united-states/.