Artificial intelligence and cybersecurity

Event report

The moderator, Mr Hussein Abul-Enein, Senior Policy Analyst, Access Partnership, explained that the session would focus on applications of AI in cybersecurity. The discussion was timely as countries have started to develop national strategies in AI, encompassing the security aspect.

Dr Jean Rickli, Head of Global Risk and Resilience at Geneva Centre for Security Policy, mentioned that AI is transforming ways of waging war, together with other technologies such as quantum computing and synthetic biology. The price of technology is falling, so more actors have access to these technologies, raising security concerns. Rickli also called attention to the geographical distribution of these technologies, which are concentrated in only a few countries.

Mr Henrik Kiertzner, Cybersecurity Division, SAS, believes there is still a hazy understanding of what AI is. According to him, AI does not yet exist, but societies are experiencing the augmentation of their capacity to perform analytics, and undergoing a revolution when it comes to digitalization. Digitalization has a significant impact on security. For many years the military thought they could cope with the pace of change, but the threats environment is much more dynamic than the defence environment.

Kiertzner questioned whether AI techniques will cope with current information overload, especially in cybersecurity. Enterprises using information systems have lost control and visibility of their networks. With AI, understanding and interpretation of what happens in an organisation can be regained, and this is a very positive development. At the same time, the use of technology against information systems is increasingly sophisticated. Cybersecurity operators can identify and manage risk, and minimize damage, but not avoid this.

Mr Rickli called attention to the use of AI to produce fake news, including images. Mr Kiertzner agreed, and added that identity and trust will be key issues in the twenty first century. Governments are falling behind the curve, and don’t have the ability to plan more than five years ahead.

Mr Antonio Amendola, Executive Director of International External Affairs, AT&T, mentioned best practices and collaboration initiatives between public and private sectors. Both sides should take a leading role in encouraging this cross-sectoral dialogue. For this to work, the private sector needs to be assured that the information they share with the public sector will not be misused. In addition, governments should also take a light approach when it comes to the regulation of new technologies, including AI. Companies need to be free to innovate and collaborate. If regulation is too strong, companies shift resources from innovation to compliance. Mr Kiertzner disagreed with this perspective, since governments have the responsibility to protect their citizens, and they do that in many different ways, from regulation to military action. Mr Rickli agreed that there needs to be regulation and safeguards introduced in innovation processes, together with a ‘kill switch’ if things go wrong.

Questions from the floor concerned how developing countries can convince the private sector to invest in AI, how to strengthen cybersecurity, and whether AI is predominantly good or bad.

Mr Amendola called attention to the importance of capacity building, of understanding the array of available solutions when it comes to cybersecurity, and the need for governments to participate in international meetings in this area.

By Marilia Maciel