Any new technology brings potential harm and disruption to society. It is important to understand the implications and consequences of Artificial intelligence (“AI”), particularly as its use becomes increasingly widespread. AI encompasses machine learning, deep learning, big data, open data, algorithms, robots and much more. Although the use of AI comes with plenty of benefits, if combined with a lack of protections, safeguards and regulations, it could impact negatively on people’s livelihoods, with potentially disastrous consequences.
Bias in AI
In the same way that bias manifests itself in humans, AI too can be biased on many different levels. The bias within AI is typically reflective of the bias of the human being who designed it. An algorithm, at its simplest level, is a sequence of instructions to solve a specific problem. Developers of AI systems can introduce bias by setting certain parameters and instructions when designing a particular algorithm. For instance, if developers choose income as a parameter in their algorithm, this may well result in an outcome which creates a racial bias (though unintended) which would prejudice a particular minority group.
AI and the immigration system
The use of algorithms within the UK immigration system are already widespread. From 2015, the Home Office applied algorithms to filter applications by establishing whether an application had a low, medium or high risk of being fraudulent. The applications identified as high risk were then passed through more rigorous checks and were thus more likely to be refused. The use of this algorithm was successfully challenged, as it was argued that the technology brought bias and racial discrimination into the immigration system. This algorithm is to be scrapped and in line to be redesigned.
Applications such as those under the EU Settlement Scheme, utilise data-sharing between government departments, and have an automated decision making system when processing applications. The benefits of such systems is increased efficiency, reduced costs, as well as quicker processing times. Nevertheless, it remains imperative that safeguards are put in place to mitigate any risks of discrimination and bias when utilising such technologies in decision-making, to ensure applications are processed fairly.
Is there a solution?
As technologies continue to expand at unprecedented levels, the importance of having safeguards in place becomes increasingly important. Over time, it may appear that the rapid expansion of technology could outpace the existing legal frameworks in place to safeguard against and challenge any potential misuse of AI.
Data protection laws, implemented by national governments, are an example of legal frameworks which aim to protect against the potential harm which could be caused by AI systems and provide mechanisms for managing the risk of their misuse. An example of this is the European Union’s General Data Protection Regulation (“GDPR”). The GDPR sets out regulations on data protection and privacy, providing individuals with control over the use of their personal data.
It has become essential not only to have in place laws to protect and safeguard human rights, but to also have competent bodies overseeing the enforcement of such laws, whilst keeping up-to-date with the advances in AI technology. Public bodies, such as the Home Office, should ensure greater transparency of their automated decision-making systems, considering the impact algorithms can have on visa applications and thus, directly, on the futures of the families and individuals submitting them. Although technological advances such as algorithms implemented by AI have allowed more efficient, accessible and cost effective systems to be developed, stringent privacy and data protection laws are crucial in AI decision-making processes.
Gherson has a wealth of experience in dealing with all aspects of UK immigration. If you have any specific questions or queries in respect of your particular circumstances, please do not hesitate to contact us.
The information in this blog is for general information purposes only and does not purport to be comprehensive or to provide legal advice. Whilst every effort is made to ensure the information and law is current as of the date of publication it should be stressed that, due to the passage of time, this does not necessarily reflect the present legal position. Gherson accepts no responsibility for loss which may arise from accessing or reliance on information contained in this blog. For formal advice on the current law please don’t hesitate to contact Gherson. Legal advice is only provided pursuant to a written agreement, identified as such, and signed by the client and by or on behalf of Gherson.