“Cyber Technologies, Artificial Intelligence and International Security”
The Arms Control & Disarmament Centre (ACDC) at the Institute of Strategic Studies Islamabad (ISSI) organised a Two Day Online Workshop on “Cyber Technologies, Artificial Intelligence and International Security” in collaboration with the United Nations Institute for Disarmament Research (UNIDIR), Geneva on January 24 -25. The theme of the second day of the workshop was “Artificial Intelligence and Autonomy.”
The second day was opened with remarks by Dr Cécile Aptel, Deputy Director, UNIDIR, who said that Lethal Autonomous Weapons Systems (LAWS) are at the heart of the discussion at the UN Group of Governmental Experts (GGE). It is important to discuss the legal and ethical implications of developments in the field of LAWS. She highlighted the technical complexities that states have to grapple with LAWS and autonomy, the importance of discussing the dual-use nature of the technology and its implications for international security.
Session I on “Artificial Intelligence and Autonomy 101” was moderated by Ms Ioana Puscas, Researcher, UNIDIR. Speakers included Dr Yasar Ayaz, CPD/Chairman at National Center of Artificial Intelligence (NCAI) & Professor of AI & Robotics at NUST, Islamabad, and Dr Giacomo Persi Paoli, Head of Programme, UNIDIR. Mr Paoli explained key concepts involved in Artificial Intelligence (AI) and autonomy. AI systems range from deterministic that are fully predictable to Non-Deterministic that are less predictable. Talking about the role of humans in AI he said that it ranges from full direct control i.e. no autonomy, to humans in the loop where humans need to validate and humans in the loop where humans intervene if necessary. Dr Ayaz spoke on how can AI be used for the benefit of humans. It is already been used widely in finance, judiciary and medical science in Pakistan. He underscored how it is a transformative technology with immense economic benefits. AI software market revenue worldwide was over US$34 billion in 2021 and is expected to increase. Countries are investing hundreds of billions of dollars in AI.
Session II on “AI, Autonomy and UN” was moderated by Mr Usman Jadoon, Director General, UN Division, Ministry of Foreign Affairs, Pakistan. Speakers included Ms Aamna Rafiq, Research Associate, ACDC-ISSI, and Ms Ioana Puscas, Researcher, UNIDIR. Ms Puscas said that discussions started on LAWS at the UN in 2014 and 11 guiding principles were adopted in 2019. She identified two distinct positions that have emerged at the UN – one that endorses the need to develop specific law in the contest of Autonomous Weapons Systems (AWS) and the second position endorses non-legally binding instruments. While there is agreement on the need for compliance with international humanitarian law, no consensus was reached over substantive matters of the 2021 GGE meeting. Ms Rafiq discussed the way forward based on the 11 principles that have been adopted at GGE. She highlighted the need for implementation mechanisms and proposals for implementation at the UN like the national legal reviews and establishing links between national and international regulations.
Session III on “Black Box Unlocked: Predictability and Understanding of AI Systems” was moderated by Dr Giacomo Persi Paoli, Head of Programme, UNIDIR. Speakers were Major General (Retd.) Ausaf Ali, Advisor Strategic Plans Division, Pakistan, and Dr Pascale Fung, Director, Centre for Artificial Intelligence Research, Professor, Hong Kong University of Science and Technology. Maj. Gen. Ali in his presentation stated that predictability and understandability are widely held to be the vital qualities of AI systems and represents an important point among the many different parties to the debate on emerging technologies in the area of LAWS and other forms of military AI. He also raised concerns regarding the AI system including extensive use in military applications, absence of legal framework, open-source access to AI system and third part risk amongst others. Dr Fung spoke on natural language processing in terms of AI systems and modern AI. She was of the view that are multiple issues that can be seen with the use and emergence of AI technologies including superior privacy, safety, sustainability and environment, legality, fairness and source bias amongst others. She shed light on the common ethical principles that are common between the Chinese government and the EU and around the world like fairness and justice, privacy, safety, interoperability, diversity, environmental wellbeing etc. During her presentation, she shed light on the foundational models and is trained for the existing text on the internet with parameters called language models which are then used for various other tasks.
Session IV on “Data and Autonomous Systems” was moderator Dr Rabia Akhtar, Director CSSPR, Lahore. Speakers included Air Cadre (Retd) Khalid Banuri, Former Director General, Arms Control and disarmament Affairs, Strategic Plans Division, Pakistan, and Mr Arthur Holland Michel, Senior Fellow, Carnegie Council for Ethics in International Affairs. Mr Holland during his presentation was of the view that all the autonomous systems and machines that operate without humans at any given stage of operation require data. However, a machine supersedes the environment and act appropriately according to the data and at times can create problems with the data that the autonomous system collected. He reiterated that machine learning relies highly on the knowledge which is developed through a process. Air Cadre (Retd) Banuri while making his remarks stated that today’s world is increasingly becoming a complex world both politically and technologically. While shedding light on the hybrid military operations, Air Cadre (R) Banuri opined that hybrid military operations suggest that a lot of teamwork and interactions require a lot of human aspects, a key element missing in machine operating systems.
The closing session of Day II was moderator by Malik Qasim Mustafa, Director ACDC-ISSI. In closing remarks Dr Giacomo Persi Paoli, Head of Programme UNIDIR while concluding the two-day workshop proceedings stressed the importance and significance of the issues of cyber technology and AI. He made a note of thanks to the Government of Pakistan and ACDC at ISSI for conducting the workshop. In his concluding remarks Ambassador Khalid Mahmood, Chairman BoG, ISSI stated that it has been a learning experience. He stated that what is important is to improve human involvement in the larger system and make sure the use of ethical grounds on issues dealing with technologies and cyberspace. There is a need to put in place legally binding instruments along with confidence and capacity-building measures. However, this requires political will and constructive engagement.Ambassador Zaman Mehdi, Deputy Permanent Representative of the Permanent Mission of Pakistan to the UN, Geneva expressed gratitude to ISSI and the team at UNIDIR while making his Note of thanks. He stated that the workshop led to fruitful discussion and has set the discourse of debate on a pressing issue.