Artificial Intelligence and Big Data – An Uncharted Territory for Migration Studies?

11.04.2017 , in ((Politics)) , ((2 Comments))

The technological advances in machine learning, big data, and artificial intelligence – for a long time confined to the realm of science fiction – are increasingly entering our daily lives. By scientifically predicting future migration patterns, changing the underlying economics of migration, facilitating legal procedures, and upgrading risk profiling and border protection, they are also highly relevant to the field of migration and mobility.

Predicting Migration Patterns

Predicting future migration patterns can be a tremendously beneficial feat. Most importantly, it will allow destination countries to make all necessary preparations for the arrival of large groups of people well in advance. A system able to predict future mass migration crises in sub-Saharan Africa is currently being tested and planned to be released as open-source software (see also the TEDx talk by Babusi Nyoni, CapeTown). Artificial intelligence is already now used to retroactively predict past migrations, such as migration during prehistoric times.

“Humans Need Not Apply”

Automatization will also affect the economics that lead to migration. Robotics might make cheap native and migrant labor obsolete. With the latest advances, however, artificial intelligence is likely to not only replace jobs at the lower end of the pay scale, but also more managerial, “white-collar” jobs. A 2013 study found 47% of total (U.S.) employment at risk. The estimates might vary, but the general trend is clear. Labor migration, with the exception of certain very highly skilled professions, could soon be a thing of the past. Sooner or later we might face the situation that humans will be subject to immigration laws, whereas non-human systems able to perform the same tasks as humans will only be subject to certain product certification requirements, but not to migration restrictions. Likewise, production-relocating policies of “bringing back our jobs” – as advocated by the current U.S. Administration mostly vis-à-vis Mexico and China – will not have the intended effect. Last December’s coup, for example, of convincing U.S. air-condition manufacturer Carrier not to produce in Mexico was a PR stunt at most: the company soon after announced that it will invest large sums in automating its Indiana factory, eventually cutting jobs anyway.

A solution to mass redundancy, which is brought forward frequently, is the Universal Basic Income (UBI). However, once individual countries introduce such a free, state-provided stipend, an even stronger political pressure against immigration might be generated, as migrants will be accused of relocating simply for the social welfare benefits.

Simplifying the Process

The legal profession is currently facing a new development, namely the advent of “legal tech”, that is, the use of sophisticated software to provide legal services. These innovations are usually provided by start-up companies such as Ross, Rechtwijzer, and others. Immigration law is slowly but surely also integrating systems that facilitate legal transactions., for example, is a privately-run chat bot that helps with the Canadian immigration process – and thus, competition for human lawyers.

Systematic analysis of big data sets as they are available to law enforcement agencies and migration authorities will at one point become a part of the migration process. Based on the information provided during an application for a visa or a residence permit, the authorities will be able to construct a detailed risk profile of the applicant. Using data analysis, it will be possible to predict the likelihood of a person to overstay their visa, to become dependent on welfare, or to commit a crime. Nevertheless, risk profiles that appear to be highly personalized might still mask systematic discrimination against groups of people, which could be enshrined in the whole setup.

It might seem like science fiction that immigration procedures could be performed by computers instead of stern immigration officials. Yet, the Canadian government started working on such a tool in 2014: the predictive analytics system of Canada’s Immigration Department shall one day provide evaluations of immigration applications. The system analyzes past decisions, including their outcome, and learns from them. At the moment, it only assists human officials, but it is quite clear where it is headed. Soon, decisions will be made primarily by the computer with only minimal supervision by humans. This will of course speed up the whole process tremendously. As a matter of fact, Hong Kong, which handles its immigration matters independently from China, has been working on such a system already since 2007!

“We Will Bring Back Our Borders”

With “The Wall” to Mexico and the various old and new barbed wire fences, walls, and border checks in Europe, border security is currently experiencing a revival. Autonomous systems will undoubtedly play a role in it. In a proposal by futurist and 2016 U.S. presidential candidate Zoltan Istvan, a swarm of smart drones could surveil the border and even interact with the people it comes across – check passports, use facial recognition to cross-check with terror lists, but also provide basic medical assistance (e.g. water) to migrants. That same proposal, however, goes way further than to just replace border guards by drones. It advocates for very authoritarian measures such as tracking migrants (particularly refugees from the Middle East) for years after they entered the country and until they have proven not to be a danger to society. In any case, it is thus a political decision where the line will be drawn – and it is a decision that will have to be made soon. Many legal and ethical questions arise: not least the question whether such systems should ever be allowed to autonomously take potentially lethal action (humans make mistakes – what if machines are less error-prone?).

Artificial intelligence brings opportunities to improve our livelihoods but also many challenges – and some would say dangers. Much is still pure speculation. But as innovation is accelerating, it is about time to address all these aspects in the specific context of migration and mobility studies.

Philip Hanke
PostDoc, nccr – on the move, University of Bern

Share on FacebookShare on Google+Tweet about this on TwitterShare on LinkedInEmail this to someonePrint this page


Chanzo Greenidge 12.04.2017

Another interesting application of big data to mobility studies is the ability to better track and measure the behaviour of diasporas. Such work would deepen conversations around remittances and also relate to networking, marketing and communications services related to diaspora outreach.

I also think we should refrain from terming technological changes as ‘advances’ if they are in fact contributing to the deepening of social and economic crises we currently face.

Reply to this comment

Didier _ 18.04.2017

There’s a flaw in the described approach to predict future migration flows in that it assumes that a model that fits past data well will necessarily make good predictions for the future. Whether the model is based on an algorithm (artificial intelligence may sound better), mathematical formula, or qualitative interpretations, it’s still a model. With regard to the remainder of the post, I’m missing a discussion of human innovation and capacity to subvert technology.

Reply to this comment