Time to start thinking about AI regulation, says KuppingerCole

It is too early to start regulating artificial intelligence (AI) because there is not even a common definition or sufficient understanding of it, but it is not too early to start thinking about it and working towards a global consensus, says Karsten Kinast, legal expert and fellow analyst at KuppingerCole.

“If we are to ensure that AI is able to bring the maximum benefit to society, we need to ensure that regulation is introduced as soon as practically possible and before it is too late, otherwise we run the risk of ending up with a mixture of ineffective, disjointed rules,” he told Computer Weekly.

The first step, which everyone in society should be working on now, said Kinast, is to start to understand the topic of AI, how it will fit into our working and private lives and what the implications of that could be. 

“We need to evaluate where we are now and where we are likely to be in the near future and the possible range of situations we can find ourselves in before we can begin to regulate it,” he said.

“This is something that has to be put on the agenda for public discussion at all levels and in all branches of society so that we can reach a consensus on what we would like to happen, so we avoid the situation that has arisen with mobile technologies, where they regulate our lives.”

Many people have become dependent on mobile devices and would not think of leaving home without them, said Kinast. “We no longer remember telephone numbers the way we did in the past because there is no need, but no one really made a conscious decision to stop doing that – it just happened,” he said.

The potential for AI to influence the lives of humans is far greater than for mobile devices, said Kinast, and that is why it is important to reach consensus around the goals and potential impact of AI to ensure a coherent and coordinated approach to regulation aimed at achieving those goals.

“This needs to be approached in a global and unemotional way so that we can come up with a set of common basic rules on AI to be followed up later by more detailed rules as our understanding grows,” he said.

Kinast believes this is important not only to ensure that the positive goals of AI are reached, but also to ensure that no single technology company or small set of tech companies is allowed to dominate setting the rules and the agenda for AI.

Now is the time to start preparing to move to regulation of AI, said Kinast, before AI that is capable of truly learning is a reality and the technology is capable of setting its own rules or modifying existing rules, instead of following rules agreed and set by human society.

The European Commission has made a start by publishing a set of principles around the ethical application of AI and has announced a major pilot in which the guidelines will be applied, but that is only one small part of the bigger picture, said Kinast. “More needs to be done and on a global scale,” he added.

“Once truly self-learning AI is a reality, it will be too late to start trying to regulate it, but in the meantime, we need to guard against AI regulating us in the same way technology is already regulating how we communicate, how we think, how we look for solutions to problems and how we interact with others.”

There will come a time in the not-too-distant future when society will have the knowledge and opportunity to regulate AI to set the rules to achieve common goals, said Kinast. “We will need to regulate AI before AI begins regulating us and influencing our decisions and actions,” he said.

Kinast will discuss this topic in more detail in a presentation entitled The global race for AI – is it time to regulate now? as well as in a panel discussion on AI regulation at the European Identity & Cloud Conference 2019 from 14 to 17 May in Munich.