The Dark Side of AI: What are the risks to the Immigration Sector?

Alchemy discusses with Kent O’Neil, Knowledge Management Specialist at Newland Chase:

No longer a sci-fi fantasy…

In past decades artificial intelligence was a sci-fi dream of the future, these days AI is a functional and ever-increasing aspect of many industries, including the Immigration sector. We are often hearing about the advantages of AI as a saver of time and money, but have we considered the dark side of artificial intelligence and its more sinister impacts on this global profession?

Alchemy Global Talent Solutions has discussed the risks and issues that AI presents to the Immigration sector with North Carolina based Knowledge Management Specialist, Mr Kent O’Neil. Kent is an expert level researcher and educator for leading Global Visa & Immigration Specialist Newland Chase, and in this capacity has become a talented prolific author on the vital subject of business Immigration.

AI’s role in Immigration today

To kick off our conversation with Kent, Alchemy expressed our curiosity regarding the current position of artificial intelligence within the Immigration sector. In response, Kent explained that ‘Thus far, the primary use has been the delivery of relatively straight-forward information on visa options and requirements to potential applicants…’

‘Rudimentary forms of AI used by governments and online visa providers interact with applicants in online visa applications systems and online service ordering – asking general questions of users and then using algorithms to give basic case assessment, suggest applicable visa options, and list process requirements. The recent addition of human-seeming “chatbots” is simply the next iteration of that technology.’

Where are we heading?

Some Immigration professionals may breathe a sigh of relief upon hearing that currently AI’s involvement in the sector is fairly limited to simple functions like automated chats – however this will not be the case for long. Kent pointedly acknowledges the efforts of several firms focused on the opportunities presented ‘by a field ripe for automation’. Kent told Alchemy:

‘I know of least a dozen online computer programs or applications that have been introduced in the last two years by private companies and law firms that purport to use AI for everything from simply answering questions about Immigration and basic case assessment to automatically completing Immigration applications and scheduling appointments with government agencies and attorneys.’

Future Predictions

So, it seems that further developments in AI for Immigration purposes are very much on the way…But what are the main tasks or functions that AI could take on in future? Kent gave us his predictions ‘The clearest beneficial application of AI in Immigration is in initial case assessment and information gathering. Immigration providers can use technology to collect initial information, assess options, provide general guidance, and give fee estimates.’

‘I could envision the technology progressing to the point that a sophisticated AI program could not only assess a case and select the appropriate Immigration route, but even start the process of obtaining necessary supporting documents once a client agrees to proceed with a case – i.e. birth and marriage certificates… a tedious and time-consuming aspect of any Immigration process.’

‘Within large multinational companies, AI and algorithms could possibly be used to initially identify and assess potential candidates for international assignments on a myriad of factors to determine which employees within the organization might have a higher potential for success as an expat – possibly even matching certain employees to the destination and assignment.’

AI risks to Politics and Immigration policy

Immigration policy and trends are often topics for political focus. Could artificial intelligence influence how global governments perceive and process Immigration? Kent explains ‘I recently read an article discussing how academics are already using similar AI applications to collect, synthesize, study, and predict Immigration trends in Africa in order to advise Immigration policy-makers.’

This is all well and good when the technology works correctly, but what if the data presented by AI sources is corrupted by outside influence? The resulting change to a Country’s Immigration policy and even attitudes towards immigrants may create a rippling effect on an international scale. If the outside influence is malevolent (for example political hackers) the results could be catastrophic – adequate security is clearly needed.

AI in the context of “extreme vetting”

Kent builds further on this point by stating ‘the greatest risk in an increasing reliance on AI and purely data-driven decision-making is if we begin to remove the intangible human quality from Immigration – understanding in an empathic sense, rather than a merely logical sense. One example of this potential negative would be, in the United States, the Trump administration’s current plans…’

Kent further explains that presently, the US government has plans ‘…to use AI in “extreme vetting” to essentially data-mine social media and other online sources using automated algorithms to identify visa holders or applicants that will become “positively contributing members of society” or “contribute to the national interest”’.

‘Even assuming the information discovered online regarding a visa holder or applicant is accurate or reveals the individual’s true nature, who is to say what is a “positive contribution” or “in the national interest? We will apparently soon find out, as U.S. Immigration and Custom Enforcement (ICE) has already been inviting proposals from technology firms to create such a system.’

What are the other risks presented by AI?

Kent’s response highlights what will be many readers’ initial concern regarding AI: ‘There are obviously the inherent risks of technological breakdowns and “glitches”. On the other hand, one could argue that humans are in inherently subject to their own breakdowns and glitches…’ This is true, so how could we resolve this Kent? ‘Optimally, AI should partner with, rather than replace, the human factor. The two should complement and provide a “check and balance” of one another.’

Kent focuses heavily on this point: ‘The speed and capacity of AI can provide greater accuracy and access, while human intelligence can still insert a necessary degree of judgment and emotion into the process… AI doesn’t replace human judgment, rather aids it with more information.’ Kent’s point regarding the need for human judgement is well noted; in an industry as ‘human focused’ and international as Immigration, could AI cope with all of the various cultural nuances involved?

‘The Human Factor’

Many readers will be thinking ‘surely AI could never reach the emotional and intellectual capacity of a human being. Humans can provide the support, subtlety or discretion required to cope with the translation between various cultures – AI on the other hand does not possess this vital ‘human factor’… Alchemy’s answer to those reader’s is ‘don’t be so sure’!

Kent sums this up very well ‘It would certainly be naïve to believe that AI will never be capable of successfully reading and responding to such complex “human” nuances as culture, social norms, religion, international relations, politics, etc. Obviously, much of the technology that we take for granted today was at a point in the not-so-distant past a reality that some would have thought impossible.’

Artificial Humanity

‘At its most basic, culture is a complex series of preferences – or “yes’s” and “no’s”, i.e. “1s” and “0s”in binary computer code – in relation to a vast set of factors and situations. Theoretically the speed and capacity of modern computers could synthesize such data into an intellectual “understanding” of culture and apply that to some degree in decision-making in the field of Immigration.’

With this additional theory provided by Kent, many Immigration professionals may be hit with the realisation that even humanity may one day become replaceable. Despite this, the reader may be soothed by the fact that the technology which may breathe ‘life’ into this sci-fi idea is seemingly ‘lightyears’ away. More soothing still, is the fact that Kent, for one, still maintains his faith in humanity…

Striking the balance

‘I believe there is a human factor which can never be duplicated – an appreciation, a respect, an empathy… for culture and people. Those aspects are, I believe, uniquely human, and to the extent that those are removed from the Immigration process, we all lose something incredibly valuable. AI can assist but can never replace that human factor.’

We hope that Kent is correct; as an agency committed to offering a talent service to individuals within diverse and multi-cultural environments, including those professionals within the Immigration sector, we feel it is important to strive towards the ideal balance between preserving the human empathy in our practice, with technological innovation.

Katie Smith, Assoc CIPD, BA Hons, HR Advisor at Alchemy Global Talent Solutions