Reported by Lee Wei Ling
On 25th of June 2021, RWC Community Series organised a Webinar titled “Artificial Intelligence and the Law” via Facebook live. It was part of RWC’s new webinar series related to Regulation Technology. The speaker was Dr Lau Pin Lean, a lecturer in Bio-Law at Brunel University in London and an active researcher for Centre of Artificial Intelligence: Social and Digital innovation. The session was moderated by Richard Wee.
Artificial Intelligence : An Overview
Dr Lau clarifies that the discussion topic on Artificial Intelligence (AI) is not about AI robots that are taking over our world but in relation to the AI system and its data. For instance, using Streamyard for Facebook live, the application of Apple Siri, a pop-out purchase notification from an app and online banking transactions are examples of AI in our daily life. She defines AI broadly as a machine that can copy human cognitive abilities such as problem solving skills and the ability of learning. Dr Lau further explains that the term “Artificial Intelligence (AI)” in most cases is commonly referred to ‘Weak AI (also known as Narrow AI)’. It means that human beings are still required to programme the machine so that it could learn from past patterns to create new information, solve complex problems, or to extract a certain kind of pattern. In short, it is a system that is designed to perform tasks more efficiently than human beings can do. In contrast to Weak AI, Artificial General Intelligence (AGI) refers to situations when machines become smarter than human beings.
To keep the ball rolling, Mr. Richard quoted the movie “The Imitation Game” which he recently watched on Netflix. This movie is projecting that the English mathematical genius, Alan Turing tries to break the German Enigma code by creating a machine. Dr Lau further expressed that Alan Turing was known as the ‘“father of AI” as he found a way to artificially assist human beings to calculate what our brain can calculate but unable to calculate in depth. Therefore, algorithms, machine learning and deep learning allow the machine to learn from data, give certain kinds of output and to avoid human errors.
How can we regulate AI?
Breaching personal data and privacy is one of the biggest challenges in regulating AI. This is because users are not aware of what they are agreeing or consenting as end users. Regulations such as EU General Data Protection Regulation (GDPR) and Malaysia’s Personal Data Protection Act 2010 (“PDPA”) are important in playing parts as regulations of AI systems or systems that employ the use of AI. Users shall be given the opportunity to opt in or opt out in giving consent when data is being used in certain ways.
Besides, data privacy also concerns surveillance, especially in tracking and monitoring. For instance, China and Russia implement AI based technology for mass surveillance purposes. In Malaysia, we constantly use the MySejahtera application to check in and log in where we are going to, and we trust that necessary precautions are taken to protect our personal data. This is helpful in contact tracing in events where an individual has close contact with a COVID-19 patient. AI has also developed in the MySejahtera application by assisting the authority to detect a new cluster as well as to keep track of the diseases and to prevent certain outbreaks.
Recently, in June 2021, the EU Commission has put forward a proposal to harmonise EU level directives and legislation to take into account AI. In 2018, the committees of the legal affairs in the EU Parliament had actually put forward a report about how the AI system should be governed. The said report particularly dealt with the legal status or ‘personhood’ status of AI systems, and whether the manufacturer can escape from some sort of liability if a legal person is imputed on the AI system. The proposed regulation in AI is a huge framework that is still going through the long legislative process. This all-encompassing is predicted to be introduced in the upcoming year of 2024 or 2025. The new EU legislation on AI will cover provisions on what kind of AI system is prohibited, transparency and government enforcement. There are also additional provisions that are imposed which are the “High Risk AI System”.
Dr Lau also shared that the European AI Board will be established to coordinate and oversee the enforcement of AI. The AI system’s definition is very broad, where it captures everything and actors at different parts of the value chain of the AI system such as the role of providers, manufacturers, importers, distributors, users and facilitators.
With regards to the legislations in AI, China has a National framework to deal with AI. Singapore has also recently launched a second edition of the Model Artificial Intelligence Governance Framework (“Model Framework”) in January 2021. United Kingdom has different legislations that touch on AI. Malaysia has yet to have any specific regulations to regulate the application of AI. Therefore, we have to draw strength from PDPA and Malaysian Communications And Multimedia (MCMC).
As for ethics on AI, Dr Lau believes that ethics must be an integral component in development of AI systems, particularly in the early stages of development. Ethics has to be developed and law must be added at each stage of development.
Application of AI in Daily Life
There are certain areas which draw strength from AI to expand the industry which can be seen from below :-
Online shopping captures one’s attention to promote and persuade us to buy things. The AI system that has been used in e-commerce has been taught to be able to figure out what consumers like based on their preferences and in accordance to the past history of search engine. The AI system can filter out the requirements of consumer in looking for a particular brand or within the price range.
Dr Lau highlighted both protection of privacy laws as well as an individual’s own initiative and responsibility as a user. She explained that the idea of this is that every time whenever you are visiting somewhere new, one must always remember to check the data and privacy statement, and see what you are sharing at that particular platform and how your information is going to be used by that platform.
A question was posed to Dr Lau on how we can comprehensively manage our risks in being AI users, especially business owners who deal with sensitive data. Dr Lau replied that business owners shall not retain information or data beyond anything that is happened to be necessary. We must make sure that consent form is available for users and compliance to the regulations of data protection. For instance, a user must be able to find clarity in the information that he/she is dealing with, if the information is going abroad or to a third party, if the data needs to be stored separately or merged in one particular server, and how are those datas being classified? Hence, business owners must protect the sensitive data of the consumer and not abuse it.
Besides, AI system can also be found in the gaming or esports sector. Dr Lau expressed that the Starcraft game is a huge victory for AI and the first of its kind. The game of Starcraft is a hugely complex military science fiction style game which requires a lot of quick thinking and making decisions promptly. The goal is to sharpen AI training, and the methods of reinforcement learning. Dr Lau explained how AI works in Starcraft game is that when a player takes something that he/she found out to create a system that can operate in an extremely complex virtual environment. The hope is that this particular development can be used for other means of developing AI. For instance, in the field of healthcare, esports, agriculture or technology in businesses. Through playing esports, the players are trained in having innovative mindsets in order to improve and make certain areas better as well as to apply them to the current practical applications. As a result, businesses, organisations and people can leverage AI to optimise their processors and enhance their work flow.
Robotics also has been in our daily life especially in the field of manufacturing, medical and healthcare where robotic surgery is performed, and robotics for the purpose of companionship. Dr Lau raised the issue of legal personality for robots for us to ponder. She enlightened us with an interesting fact that Sophia is given citizenship in Saudi Arabia as the first robot. The development of AI in Sophia allows her to hold eye contact, recognize faces, understand human speech, and express human expression. However, to what extent she will be able to exercise her citizenship and her rights under the constitution is being questioned.
Conclusion
In a nutshell, we cannot avoid the way technology has affected our life but we can learn how to be a responsible and careful user. Dr. Lau also encourages people to start practicing to read the fine print. We must not think of AI as a big takeover in our life despite that takeover has occurred in several industries. Instead, we should think about how technology can be innovative for us by helping us and enhancing ourselves to deal with problem solving, reduce our administrative burden, and carry out due diligence work.
Published on 8 July 2021