2022.Nov.14

REPORTS

The 2nd BAIRAL Research Meeting for 2022 Report on “International Politics and AI: From Cybersecurity to Killer Robots”

On May 26, 2022, BAIRAL (a study group organized by research assistants of the B’AI Global Forum) held its second meeting for 2022 online. The guest speaker was Mr. Hitoshi Sato, who is a doctoral student in the Graduate School of Social Design Studies at Rikkyo University. In the meeting, hegave a presentation on the current situation of military use of AI and regulations on killer robots around the world, including issues regarding the use of AI in the military field, which was followed by a discussion on related topics with the audience.

Regarding the discussion surrounding the use of AI in the military, Mr. Sato said that while some view all military use of AI as bad, there is also an opinion that AI-equipped robots are better suited to many military tasks than human soldiers. For example, there are 4D tasks: Dangerous (bomb disposal), Dirty (handling hazardous materials), Dull (border security), and Distance (reconnaissance and offense in places where humans cannot go, such as underground and underwater). He also mentioned some aspects that the use of robots can ensure “human security” for military personnel, and the aspect that AI recognition is often more accurate than remote control by human military operators.

On the other hand, as for concerns about the military use of AI, he introduced the example of a killer robot “LAWS” (Lethal Autonomous Weapons Systems). He said that LAWS has issues that there is no clear definition of it and the actual reality is not known, including to what extent robots will be autonomous and make decisions, and human intervention will be involved. Another concern is the risk that they could be converted and developed into autonomous robots by cyber-attacks, regardless of the intentions of the government or manufacturer. Additionally, there are also concerns that LAWS could autonomously attack and kill certain ethnic groups, using biometric-based facial recognition technology to detect them.

Finally, Mr. Sato talked about the discussion and the current status of regulations that could restrict killer robots. For example, the “Protocol on Blinding Laser Weapons,” an international treaty that totally bans the use and transfer of laser weapons designed to cause permanent blindness, is a reference for LAWS. He also noted that it may be possible to restrict LAWS through the regulation of Weapons of Mass Destruction, international humanitarian law, etc., but that no progress has yet been made in such regulations.

After his presentation, the participants discussed the regulation of killer robots, pointing out the aspect that it is difficult to obtain the image of killer robots because the actual utilization of them is unknown, compared to the dangers of drones for surveillance and reconnaissance, and the need to separate the regulation of their use from that of their development.