Immigration laws, policies, and practices are challenging the ability of many communities, including the artificial intelligence and machine learning (AI/ML) community, to incorporate diverse voices in their work. As a global, multi-stakeholder non profit committed to the creation and dissemination of best practices in artificial intelligence, the Partnership on AI (PAI) is uniquely positioned to address the impacts of immigration laws, policies, and practices on the AI/ML community.
PAI believes that bringing together experts from countries around the world that represent different cultures, socio-economic experiences, backgrounds, and perspectives is essential for AI/ML to flourish and help create the future we desire. In order to fulfill their talent goals and host conferences of international caliber, countries around the world will need to devise laws, policies, and practices that enable people around the world to contribute to these conversations.
Based on input from PAI Partners, and PAI’s own research, this paper offers recommendations to address these specific challenges. It highlights the importance of conferences and convenings for a variety of disciplines that are making important contributions to AI/ML, and makes recommendations for participants and organizers that may facilitate ease of travel for these events. It also presents recommendations for governments to improve the accessibility, evaluation and processing of visas for all types of potential visitors, including students, interns, and accompanying families. Appendices to the paper respond to potential questions, and provide an overview of the global demand for AI talent, as well as additional details on technical or expert visa, residence and work permit laws, policies and practices.
PAI’s recommendations are based on our area of expertise, and have been developed to help advance the mobility of innovative global AI/ML talent from a variety of disciplines. Many countries have already created visa classifications for other specialized occupations, including medical professionals, professional athletes, entertainers, religious workers, and entrepreneurs.
At the same time, we acknowledge the complex immigration debates taking place in countries around the world, and the challenges posed by global migration and the quest for basic human rights and dignity. These recommendations are in no way intended to minimize or replace opportunities for those affected by the ongoing immigration discussions and policymakers actions. We hope policymakers can create a path towards permanent residency or citizenship for these groups. In fact, while our recommendations target our field of expertise, we hope our paper can serve as a useful resource for the broader community, in support of balancing government public safety responsibilities with the benefits of immigration, freedom of movement, and collaboration.
Though this document incorporated suggestions from many of PAI’s partner organizations, it should not under any circumstances be read as representing the views of any specific member of the Partnership. Instead, it is an attempt to report the views of the artificial intelligence community as a whole.
Frequently Asked Questions
Why would PAI tackle a subject such as visas and immigration? This topic is not really related to artificial intelligence research.
PAI believes that bringing together experts from countries around the world that represent different cultures, socio-economic experiences, backgrounds, and perspectives is essential for AI/ML to flourish and help create the future we desire. Artificial intelligence is projected to affect all facets of society, and in some ways it already is having those effects. PAI’s work addresses a number of topics related to AI, such as criminal justice and labor and economy. Our work to address immigration challenges affecting the AI community is quite similar.
How does this document pertain to PAI’s mission and work?
This document makes visa policy recommendations that would improve the mobility of global AI/ML talent and enable companies, organizations and countries to benefit from their diverse perspectives. Fostering, cultivating, and preserving a culture of diversity and belonging in our work and in the people and organizations who contribute to our work is essential to our mission, and embedded in our Tenets. These include: committing to open research and dialogue on the ethical, social, economic, and legal implications of AI, ensuring that AI technologies benefit and empower as many people as possible, and striving to create a culture of cooperation, trust, and openness among AI scientists and engineers to help better achieve these goals.
Who benefits from this policy paper?
Unlike large, multinational companies and prominent, well-funded universities and colleges, startups, small- and medium-sized enterprises, individuals traveling to conferences, less affluent applicants, students, and interns often lack the resources to hire experts to ensure their preferred candidates have the greatest chance to obtain visas for internships, to study, or to work in their organizations. These groups and individuals often cannot successfully compete for visas, especially those that are numerically limited. They would be the greatest beneficiaries should governments implement these recommendations.
Why is PAI uniquely suited to address this issue?
As a multi-stakeholder non profit, PAI convenes over 100 global Partners, originating from 12 countries and four continents, and representing industry, civil society, and academic and research institutes. As such, we are uniquely qualified to describe the impacts of immigration laws, policies, and practices on the AI/ML community. The impetus for this document came from many of PAI’s Partners and colleagues, who have shared how certain visa laws, policies, and practices negatively affect their organizations’ abilities to benefit from global representatives and perspectives in their work.
Why is PAI focused on incorporating diverse voices in AI/ML?
Diverse perspectives are necessary to ensure that AI is developed in a responsible manner, thoughtfully benefiting all people in society. Voices and contributions from global talent are also essential to reducing the unintended consequences that can arise from AI/ML development and deployment, including those related to safety and security. Due to the emergent and rapidly evolving nature of AI technology, AI in particular engenders high impact AI safety and security risks, which can be mitigated by increasing the diversity of participating voices1 Diverse representation also serves to promote the safety of key members of the AI/ML community. Underrepresented voices, such as those of minorities and the LGBTQ community, are important as we design AI/ML systems to be inclusive of all populations.
Is PAI suggesting that AI/ML practitioners should be treated differently than other skilled workers? How is this different from other visa categories?
PAI’s recommendations would enable AI/ML practitioners, from a variety of disciplines, to travel and work more freely. In some cases, this could entail special visa classifications, similar to those that already exist for skilled workers in other specialized occupations, such as medical professionals, professional athletes, entertainers, religious workers, entrepreneurs, skilled laborers and trades workers.
This paper also highlights the many disciplines involved in the development and operations of AI/ML systems, above and beyond what is sometimes defined as “skilled technology work.” Responsible AI/ML systems involve input from researchers and practitioners in social sciences such as economics, sociology, philosophy, ethics, linguistics, and communications, and the “experiential expertise” offered by those working in labor and workers’ rights2, in addition to technical fields such as mathematics, statistics, computer science, data science, neuroscience, and biology.
How does this work? Unlike medical professionals or engineers, AI/ML practitioners don’t have a certificate or license for governments to determine that they are experts.
Countries establish criteria for evaluating applications, whether for technical talent, a professional athlete, or someone skilled in trades or labor. Established eligibility criteria, and the process for evaluating this criteria, vary greatly from country to country. The PAI paper offers models for countries to consider and draw upon if they decide to create a classification for AI/ML practitioners.
For example, some countries require letters from a potential employer, or to have someone in the field attest to the applicant’s particular skills, or other supporting documentation that proves the applicant has the desired skills. Some examples:
- An independent review board: The UK Tech Nation Visa, also known as the Tier 1 Exceptional Talent Visa, assigned an independent, “designated competent body,” to review and endorse applications. The Tech Nation Visa Guide outlines the skills and specialties typically exhibited in applications reviewed by this independent body, and the eligibility criteria.
- Points-based system: Canada’s Express Entry Program, like other Canadian visas, evaluates applicants on the basis of the types of occupations and levels of skills they hope to attract. Certain occupations and skills, among other criteria, garner greater numbers of points. The higher the overall point total, the greater the likelihood of being admitted entry.
- Government review: Japan’s Skilled Labor Visa program seeks documentation to support the visa application, and that documentation must prove, among other elements, that the applicant has a certain number of years of experience. The government will review the documentation, and issue a Certificate of Eligibility (COE) if they think the applicant possesses the necessary experience and skills. The existence of the COE in the application can accelerate the visa processing time.
- Additional examples can be found in Recommendations for Policymakers #1 and Appendix C of the paper.