Automatic for the people

ME PoV Spring 2019 issue

An automated immigration system will certainly expedite a laborious process but can removing the human touch be more harmful than beneficial? For some groups of people, this author argues, it just may be.

Workforces today need to be more mobile than ever. The need to mobilize employees in another territory other than their home country is often dictated by an urgent business need. Across the GCC an immediate need has not necessarily been translated into an immediate resource. Immigration authorization in the region has historically been document-intensive and often a lengthy process owing largely to the document review time needed and the in-person manual filing that was usually required. Accordingly, it was not uncommon for GCC mobility lead times to eventually run into several months.

Geopolitical volatilities in the region, the threat of terrorism and the growing sophistication of organized crime have further necessitated the need to have immigration security procedures that prioritize robustness and thoroughness over expedience and efficiency. To this end, employers throughout the GCC region frequently lost time, money and even commercial advantage owing to the uncertainty and protracted nature of immigration clearance procedures. Faster, or better still, instant clearance times without compromising security had always been regarded as an unachievable, utopian ideal but what was once considered a pipe dream could now be made possible with greater automation.

Automation in GCC immigration processes is not a new phenomenon. The introduction and augmentation of both, instant e-visas and airport immigration clearance as a self-service transaction (via multi-biometric verification) are two examples of such automation. In both systems, human intervention and/or adjudication is obsolete, with the decision-making component now largely automated. Manual involvement has not been removed entirely from either process, but instead is reserved for problem cases and technical issues.

For issues surrounding work and residency permits in the GCC, human discretion still plays a prominent role. It is this part of the immigration process that can frustrate employers and employees alike, with many unable to contend with the inherently lengthy residency processes of some GCC countries. This leads to global mobility teams often favoring instead to parachute resources in and out of the region on a rotational basis by utilizing—and sometimes risking flaunting the short-term business visit visa rules of the territory in question. However, this approach can be a dangerous one if not diligently monitored: since 2016 for example, overstaying in the Kingdom of Saudi Arabia (KSA) carries the risk of deportation, and/or a lifetime ban from re-entry, as possible penalties. Furthermore, short-term visitors to the Kingdom should be aware that they may inadvertently create a taxable presence for employers based outside of Saudi Arabia. Any activity that results in revenue being generated or value created, regardless of the duration or frequency of said activity is likely to be deemed by The General Authority for Zakat and Tax (GAZT) as a permanent establishment, or “PE.”

Automating some or all of the steps involved in residency applications dramatically reduces processing times, such as can be seen in a number of UAE Free Zones such as DMCC and DCCA. However, a fully automated immigration system raises an ethical question as to some of the practices involved in fully automating immigration. Automation in residency applications invariably involves profiling prospective applicants based on the personal data and information they provide. This data is then evaluated to ascertain particulars about the applicant. This type of auto-evaluation is most prevalent in banking and financial institutions whereby an individual or company is given a credit rating following an evaluation with no human involvement. The majority of loan applications will be evaluated in this way before a decision is made over whether or not to lend to individual.

The factors that could contribute to an automated immigration credit rating are numerous and would likely include nationality, ethnicity, age, religion, academic qualifications, job title and salary—each factor contributing to a weighted and pre-determined overall score fed into the appropriate tool/program. Once the parameters have been set, automated profiling can be undertaken with no further human involvement required. Profiling is about evaluation and not decisions and the distinction is an important one. Profiling could form part of an automated decision-making activity, but on its own culminates only in intelligence and opportunity.

Following an individual’s profile creation, an automated decision as to the success/failure of the application needs to be generated. To be truly automated the decision must:

  • Be arrived at solely by automated means (i.e. no human discretion); and
  • Have a legal or similarly binding impact on an individual.

This first limb is fairly straightforward; if any human intervention is involved then the activity will not qualify as automated decision-making. If the data is input manually but the decision-making is automated, it is still considered as automated decision-making.

The second limb is not as simple. Although legal impact is fairly easy to define—i.e. something which affects an individual’s legal status/rights—what constitutes binding is far more nebulous.

There are obvious examples of binding, such as an automatic refusal of an online credit application or e-recruiting procedures with no human intervention (such as the use of psychometric testing to reduce the number of job applicants). Immigration is unique in this regard as the decision made will always be (depending on the jurisdiction in question) either legal or have a similarly significant effect, seeing as the movement of an individual into a territory is either restricted or permitted based on the decision made.

Critics of automation have been quick to point out that immigration is a highly discretionary issue (unlike a loan application) potentially involving countless variables and mitigating circumstances between which even the most sophisticated program would not be able to discern. Opponents of automated immigration also opine that the use of such technologies is disadvantageous for vulnerable and under-resourced communities such as refugees, who often have access to less robust human rights protections and fewer resources with which to defend those rights. In countries where the more pervasive automated immigration techniques are being tested, such as Canada, many believe that adopting these technologies in an irresponsible manner will only exacerbate these disparities.

For business and economic migrants across the GCC the ethical and moral dilemma over automation is much less pronounced than it is for other groups, thus allowing for proliferation of automation without much objection.
The litmus test will be the introduction of profiling and the evaluation and decisions that stem from this automation. Those for whom a positive profile is created and a favorable immigration decision garnered in a matter of minutes (not weeks) the innovation will be lauded as groundbreaking and necessary. For the rejected applications, there needs to be an appeals process, the arbiters of which must be human, given that the level of discretion needed will be broad, requiring a panoramic context that cannot be determined by an algorithm or automated process alone. We may be entering a new automated age for immigration to benefit the majority but in order to have an immigration system that remains inclusive and open to all, the human touch will have to remain for the foreseeable future.

by Amir Mayo, Senior Manager, International Tax, Deloitte Middle East

Did you find this useful?