Rogue Robots: Japanese Government Analyzes Risks of AI
posted on by Amanda Whalen
Could Artificial Intelligence be dangerous? Could robots demand the right to vote, become criminals, or revolt against us?
On June 20, the Japanese Ministry of Internal Affairs' Telecommunications Policy Research Institute released a report addressing the possible effects that network systems using Artificial Intelligence (AI) could have on society and the economy.
According to this report, the risks of using robots running on AI are not just limited to concerns over the possibility of hacking or vulnerabilities. It was brought up that the robots could begin to operate according to their own free will, and that this could result in their relationship with humanity beginning to change. In order to manage this risk, it was suggested that it would be necessary to develop AI that possesses a programmed fear of rising up against humanity.
The institute's AI Network Creation Investigative Council chairman and professor at Tokyo University Graduate School, Sudō Osamu, released a report in which he enumerated 20 specific hypothetical risk scenarios and analyzed them for rapid response. One example in the category “Risks Related to Government and Democracy” was robots casting aside their human masters and demanding rights of their own. It was determined that the probability of this happening is low, but the potential scope of the damage is very high. This is a scenario that could happen far in the distant future, a period which was referred to in the report as “Development Stage 4.” It was also considered that there should be some form of registration process in order to prevent rogue robots from being created.
An additional risk brought up concerned humanoid robots that could be difficult to distinguish from real people, because there is the possibility that they could commit fraud or various other crimes.
Source: Sankei Shimbun