AI in migration is fuelling global inequality: How can we bridge the gap?

Artificial intelligence (AI) in migration and mobility systems has recently become a hot topic in research and policy circles, given the heightened prominence of the technology.

But as AI deployment rapidly advances through different policy domains, you might be surprised to learn that AI in migration has been present for several decades. This fact, together with major, long-term trends in international migration, points to the risk that AI technologies in migration and mobility systems are on track to exacerbate digital divides both between and within states.

AI usage for migration and mobility management was, for some countries, a logical consideration in the 1990s, given the significant, sustained increases in international air travel, the deployment of online visa application systems and growing international border crossing data capture. The use of AI centred on managing increasing volumes efficiently and enabling enhanced data capture for strategic analytics.

But of course, those countries had existing advanced data systems and the resources needed to build on existing administrative data collection toward more elaborate AI-supported systems.

Immigration authorities in Australia, Japan, the United States, many European countries and Hong Kong, China initially sought to streamline processing systems using machine learning and other technologies. The immigration department in Hong Kong, for example, built its AI e-Brain system to process applications and build procedural knowledge through machine learning, reducing the need for case officer handling and faster processing.

The prerequisite to AI uptake is information and communications technology (ICT) digital capability, particularly the digital data capture of processes and applicants’ identity data. These require access to ICT infrastructure, electricity and ICT-skilled staff, whereas many countries worldwide lack these critical necessities, especially least-developed countries (LDCs).

These access issues further widen the gap between states, adding to the digital divide and the structural disadvantage experienced by LDCs in migration management. And the “asymmetry of power” in AI for migration globally is an ongoing problem.

It is not just inequality between states that will impact migrants. The move toward greater digitalization of migration management and increased use of AI, including for visa services, border processing and identity management, will increasingly require potential migrants to be able to engage with authorities via digital channels.

There are also calls for supporting migrants with digital access once they have arrived in a destination country and are required to navigate digital service channels for integration support. But first, potential migrants must be able to access entry-related digital processes. And this poses obstacles for many people around the world.

Access to digital platforms varies significantly within countries in some parts of the world. Data on internet access by sex and country development status, for example, show that females have much lower rates of access compared with males in developing countries and LDCs.

Notwithstanding important human rights concerns, digitalization can benefit migrants and states by reducing human error and speeding up processing. However, many people will be completely locked out of such processes, especially women from developing countries.

This exclusion is very concerning given long-term trends, which show that over the last 25 years, we have witnessed growing mobility inequality, with people from developing countries increasingly unable to access international mobility pathways.

Analysis shows that most international migration activity is now taking place between wealthy countries as both origin and destination, with policy and resource impediments preventing global mobility from poorer countries.

We can also see a growing gender gap in international migration has really taken hold in recent years, with the male growth in international migrants now outstripping female growth significantly. For example, in 2000, there was almost gender parity among international migrants, with males accounting for 50.6% and females 49.4%; by 2020, the gap had grown significantly to 52.1% male and 47.9% as the result of long-term systemic change.

The expansion of digital technologies and their application in AI-supported migration systems in wealthy, highly-developed countries will further exacerbate the growing digital and mobility divides between and within states unless effective action to support developing country populations and authorities is developed and rolled out.

There have also been UN calls for tech development and investment in LDCs that would provide a way forward for tackling and stemming the growing global digital divides. That means not just concentrating efforts at the governmental level but also supporting communities to gain access to digital technology, including women in developing country contexts.

The benefits to migrants of faster, better technologies (including AI) to underpin migration are evident. But, how this is done is critically important from equality and human rights perspectives. That includes ensuring these technologies don’t inadvertently lock out communities and put further pressure on them and the authorities. Such a consideration is especially important when the international community is committed to advancing the Sustainable Development Goals and the Global Compact on Safe, Orderly and Regular Migration.

Supporting access to safe, orderly and regular migration requires active support for digital equality. Moving forward, we must recognize and support ICT access expansion with a clear eye on rising mobility inequality negatively impacting our migration futures.

The views expressed in this article are those of the author alone and do not necessarily reflect those of IOM or any other organizations with which the author is affiliated.