Killer robots are not part of a science fiction story as they could be Terminatorbut they have become vital part of conflicts in many parts of the planetfrom autonomous machines to automated drones.
Since 2014, the UN has held different assemblies of diplomats to reach a legally binding agreement to regulate and prohibit its use by 2026, due to several factors that have to do with the idiosyncrasies of the technology, ethics and the legal field.
«It’s very easy for machines to confuse human targets,» Wareham of Human Rights Watch said in a United Nations interview. «People with disabilities are at special risk because of the way they move. Their wheelchairs can be mistaken for weapons.»
And he adds more warnings: «There are also concerns that facial recognition technology and other biometric measurements cannot correctly identify people with different skin tones. AI continues to have shortcomings and carries the biases of those who programmed these systems.»
Beyond the way the technology proceeds, there are also concerns in the legal field, since if a killer robot kills a civilian or commits genocide, Who would be ultimately responsible? The manufacturer? The algorithm programmer?
In the absence of a globally binding regulatory body, many civil rights organizations have begun to raise their voices, with the aim of preventing the use of these killer robots in security, military or border control tasks.
A cry against killer robots: «Less autonomy, more humanity»
Without falling into a technophobic or cybercatastrophic attitude, there are global initiatives to stop the development of so-called killer robots, which are machines with the autonomous ability to choose targets and, consequently, kill them.
In this way, organizations such as Stop Killer Robots have appeared –Stop the Killer Robotsdue to its translation into Spanish–, whose motto is clear and concise: «Less autonomy, more humanity.» Inside, there are also Spanish associations for civil rights.
As confirmed by Nicole Van Rooijen, executive director of the organization, in the same UN interview, the use of this type of technology would imply an added difficulty to the complex international scenario, such as determine responsibilities for war crimes.
The United Nations, represented by its Secretary General, António Guterres, has been positioning itself against the use of this type of autonomous weapons for several years and warning that «the militarization of artificial intelligence represents a serious danger.»
For her part, Izumi Nakamitsu, director of the UN Office for Disarmament Affairs, is even clearer regarding the world organization’s official position on what is considered «digital dehumanization.»
«The Secretary-General has always said that using machines with fully delegated powers to make the decision to take a human life is simply morally repugnant,» he says. «It should not be allowed. In fact, it should be prohibited under international law. That is the position of the United Nations.»
The use of AI for military purposes
Google had previously proposed not using AI for military purposes, although everything changed at the beginning of this year, at which time Alphabet changed its mind to start investing in military projects.
And Mountain View’s are no exception, since others like Palantir or Microsoft also use their models for the same purposes. Even OpenAI, the company behind ChatGPT, has reached a deal with the Pentagon for $200 million to develop AI technology for military programs.
According to Stop Killer Robots, the incorporation of AI in military applications also carries significant risks, such as «the possibility of an unwanted escalation of conflicts», with the consequent increase in civilian casualties that have nothing to do with the business of war.
To this we must add the lack of accountability, in addition to the already mentioned incompatibility with international humanitarian law, as these proprietary models are quite restrictive in terms of transparency of large technology companies.
One of the most well-known and recent examples is that of Israel’s attack on the Gaza Strip using the AI of its Habsora system to supposedly identify Hamas terrorists. Although it is not an autonomous weapon, different reports concluded that there were civilian casualties.
«Furthermore, the potential reduction of individuals to data points based on specific characteristics such as ethnicity, gender, weight, gait, etc., raises serious questions about how target profiles are created and, in turn, how they are selected,» the organization concludes.
Get to know how we work in NoticiasVE.
Tags: Artificial intelligence, robots