Artificial Intelligence is one of the buzzwords of our century, and interest in the sector has risen to a substantial amount. Malta's regulatory authorities are planning to capitalise on this phenomenon by attracting AI investors to the country.

Jackie Mallia from the AI Task Force heads the Legal and Ethics Working Group and in this interview explains what Malta's future in AI has to offer in our economy.

Malta’s recently launched AI strategy aims to set a path for Malta to gain a global strategic competitive advantage in the field. The strategy is built on three pillars: investment, startups, and innovation; public sector adoption; and private sector adoption. What is the strategy’s vision for each one?

In November 2018, the Maltese government appointed the Malta.AI Taskforce to advise upon and develop a national strategy on Artificial Intelligence. The strategy was launched on October 3 and laid out what is essentially a project plan, with three vertical pillars focusing on boosting investment, innovation and adoption of

AI as well as three horizontal enablers which apply generally to support all the three aforementioned focus areas. 

The first pillar commenced a process for several incentives to be developed by government to assist AI-related entities. Through these incentives, Malta aims to attract and cultivate the right environment for the AI sector to thrive, by setting up Malta as a launch pad for the development, testing and scaling of AI products. The strategy also puts into place measures which seek to attract the right talent and private investment for this purpose. 

The second pillar looks at adoption of AI by the Maltese government, with the aim of implementing several pilot projects to provide better services to citizens and businesses and generally improving the internal operations of the public service and sector. 

The third pillar focuses on government assistance for the adoption of AI by the private sector. The three additional ‘enablers’ relate to measures in terms of law and ethics, infrastructure and education which are intended to support the strategy’s three main goals.

Apart from the opportunities which AI opens up, the strategy also focuses on potential risks. What, if any, are the real risks of AI and how has the strategy sought to address them?

People’s imaginations tend to run wild when AI is mentioned; but the reality is that the AI that exists in our world today is what we call ‘weak’ or ‘narrow AI’, which is programmed to perform a single task, obtaining information from a specific data-set. Thus, narrow AI is not sentient or conscious, nor is it capable of emotion the way that humans are. Current systems are not close to having human-like intelligence, and therefore we will not be seeing anything like Ava from Ex Machina any time soon. 

Nonetheless, machines are already able to process data more efficiently and more accurately than humans can, and this could be seen as a risk in the sense that some jobs could be displaced through automation and intelligent systems. 

Dr Jackie Mallia from the AI Task Force heads the Legal and Ethics Working Group.

Additionally, AI systems are designed to learn to accomplish specific tasks by processing large amounts of data and recognising patterns in the data and they can become relatively autonomous. Such autonomy brings about a number of moral and legal questions – for example, who is responsible for a decision taken by an AI system that goes awry? Who owns the intellectual property rights in the works created by that system? Given that AI requires colossal amounts of data to be properly trained, questions around privacy and bias also arise. AI can also potentially used for harm, rather than good, such as employing AI systems for terrorism or other forms of attack. These are just a few of the risks associated with AI systems. 

The Malta AI Strategy deals with measures for education and re-skilling of the workforce, in order to address job displacement, as well as providing for the setup of a committee on ethics as well as a law review committee that will carry out a gap analysis of our current laws and make suggestions as to where additional legislation may be required. In additional to this, the Taskforce has also put together a comprehensive ethical framework entitled ‘Malta: Towards Trustworthy AI’ which lays down a number of ethical principles that must apply to the development and use of AI in Malta, as a way of mitigating these risks. 

Finally, the strategy also caters for a national AI certification programme, which will largely be based on Malta’s Ethical AI Framework and its underlying control framework and which would provide applicants with valuable recognition in the marketplace that their AI systems have been developed in an ethically aligned, transparent and socially responsible manner. The certification programme will be managed by the Malta Digital Innovation Authority (MDIA).

AI could introduce automation for a number of jobs. What risks, and opportunities, would this present?

It is a fact that automation could displace certain jobs, but that is a somewhat pessimistic way of looking at the concept of AI. Rather, one should view AI as a tool to enhance one’s performance – much as medical imaging equipment has assisted doctors in reaching a more accurate diagnosis, or the installation of automated teller machines (ATMs) and the uptake of online banking has aided the banking sector. Mundane, repetitive jobs, across all industries, will eventually be replaced, but this should be viewed as an opportunity to perform more interesting, satisfying work in a more accurate and efficient manner.

In addition to the strategy, you have also been working on a separate ethical framework to support AI practitioners to manage potential risks ads identify opportunities to encode a higher ethical standard into AI. Has this document been finalised and what are its main principles?

I have alluded to the Maltese ethical framework ‘Malta: Towards Trustworthy AI’ above. This has indeed been finalised and is available online.

Malta’s ethical framework builds on the Ethics Guidelines for Trustworthy AI published on April 8, 2019 by the European Commission’s HLEG as well as the OECD Recommendations of the Council on Artificial Intelligence. It establishes four ethical principles for trustworthy AI, namely: human autonomy – humans interacting with AI systems must be able to keep full and effective self-determination over themselves; harm prevention – AI systems should not cause harm at any stage of their lifecycle to humans, the natural environment or other living beings, fairness – the development, deployment, use and operation of AI systems must be fair; and explicability – end-users and other members of the public should be able to understand and challenge the operation of AI systems. 

Although led by the Malta.AI task force, the formulation of the national strategy included input from multiple stakeholders and was also open to public consultation. What influence did this have on the final strategy?

In addition to the taskforce, several working groups were established for each of the pillars and enablers, the members of which were very much involved in the discussions around the strategy. With Malta aiming to be inclusive and holistic, discussions were also held with different groups of stakeholders to ensure that the policies within the strategy are not only consistent with emerging international standards and norms, but also feasible within the Maltese environment and culture. 

Each action laid out in the strategy was therefore jointly developed with relevant stakeholders, who now have responsibility to implement them under the MDIA’s supervision. Public consultation is always a fundamental aspect of any governmental strategy, and accordingly, in March 2019, the Malta.AI Taskforce published a high-level policy document, ‘Malta – Towards an AI Strategy’ which laid out the blueprint and the key areas of focus for the Malta AI Strategy. The final Malta AI Strategy is hence a result of discussions with various experts, stakeholders and the general public

In which sectors would AI fuel positive disruption?

It is difficult to pinpoint the sectors which will be most affected by AI, as AI technology is a powerful force driving innovation across all sectors. AI can encompass anything from aiding online searches, to mitigating effects of climate change, to autonomous cars and anything in between. As such, it will impact our lives in both the business and personal spheres. By way of a few simple examples, businesses can benefit from AI by identifying and addressing inefficiencies and automating systems, such as using chatbots for customer-care support; individuals can benefit by using personal assistants to automate everyday tasks. 

The possibilities of AI are endless, and it is not easy to anticipate the extent to which the technology will develop. Certainly, the impact of AI will need to be studies within each sector separately. This also applies from a regulatory point of view – it is unlikely that we will see a generic AI legal framework rather, the implications, risks and opportunities of AI will need to be addressed sector by sector.

AI technology is a powerful force driving innovation across all sectors- Jackie Mallia

What are your personal expectations of how AI will change our lives?

AI is simply a tool which is used to amplify human intelligence. With this in mind, if knowledge is power, then AI can make us more powerful as human beings. Unfortunately, as always, this is not always a positive, and that is why the necessary controls and the necessary regulations need to be in place. Used in the right manner, however, AI has the ability to improve human life in leaps and bounds. 

On a personal note, the application of AI technology in healthcare is one of my main areas of research and interest – the technology’s use in supporting clinical decisions and information management alone can take the sector to a new level in terms of speed and accuracy of diagnosis processes, treatment protocols, development of drugs including personalised medicine, patient monitoring and care, and clinical trials. My expectation is that with responsible regulation and monitoring, AI can and will lead to the general betterment of human lives on a large scale.