Securing the world against Rogue Robos

Several Movies have captured the ugly face of Technology.  Most of the time this is because technologists intoxicated with the power of technology often create monsters without knowing the consequences. With the growth of Artificial Intelligence and humanoid robots, the dangers are increasing everyday and we need to respond to this alarming situation.

A Scary incident has now been reported from a Lab in Japan. The incident reportedly occurred in August 2017 and a whistle blower has revealed it now. In this incident 29 humans were killed in a lab producing autonomous war robots. The four soldier robots went rogue and started shooting the humans. Out of these three were dis assembled physically by the workers while the fourth was smarter and was searching the satellite data base on how to re-arm itself.

Ultimately this also might have been dismantled but the fact remains that the dangerous phase of AI is now before us and if we continue to act intoxicated and donot learn from our mistakes, we are creating  robot monsters which will destroy the humanity.

But some of the videos in the link Six scary things about AI raise concern about the capacity of the robots to think on their own and also express views about conflicts with the human race make everyone sit up and take notice of the possibility that the movie kind of situation may become a reality too soon for comfort. We may not be able to find a Rajnikant in real life to save us from the damages that rogue robots may create.

It is time that the international community takes some corrective action to ensure that “Artificial Intelligence Does not over ride the “Isaac Asimov Principles of  Ethical Robotics“.

Japan was instrumental to the first atomic bomb being dropped and now it appears that it can be the source of the next great tragedy on the planet.

Imagine what could be the consequences of terrorists either acquiring these “Autonomous Military Robots” or hacking into some of them. If we donot have a security solution for such incidents, it is better we donot create these monsters.

Earlier Incidents

From January 25, 1979 when an accident at the Ford Factory claimed the first human life by a Robot to the reported death of 29 persons in a Japanese lab by a “Rogue Robot Soldier” being manufactured, there have been several accidents where Robots have claimed human lives.

Some of them are captured here.

  1. 25th January 1979: Robert Williams killed in the Ford Factory at Flat Rock, Michigan. This was dubbed as an accident since the worker ignored a safety measure and accidentally switched on the machine while trying to repair it. (Refer here).
  2.  May 7, 2016: Joshua Brown killed in a self driving car accident when his Tesla failed to distinguish a Tractor trailer in front from the sorrounding bright sky and drove under the 18 wheel trailer and crashed. This was clearly a failure of the software and caused by the negligence of the developer and deficiency of testing.
  3. 2007: Nine South African Soldiers were killed and another 14 wounded after an anti aircraft weapon (Oerlikon GDF-005) started shooting by itself. It could be termed as a techno mechanical failure where the gun jammed and exploded before going berserk and firing 250 rounds. Software failure was not ruled out.
  4. July 7, 2016: Police in Dallas used a robot to kill a dangerous killer who had killed several persons and hiding in a building by attaching a grenade to a robo and sending it to the garage where he was hiding and exploding it. The person killed was Micah Johnson but the use of the “Bomb Detecting Robot” to execute the human was perhaps a justified action of the police under the circumstances. (P.S: It would be interesting to know how the pseudo human rights activists and the Indian Judiciary would react if such action is taken in Kashmir by the Military)
  5. 9th December 1981: An accident at Kawasaki heavy industries killed Kenji Uranda, a 37 year old man who was trapped by the working arm of the robot when it was being repaired. (Refer here). This also can be identified as an accident caused by the negligence of the worker.
  6. 2015: A man was reportedly killed in Baunatal, Germany in the Volkswagen plant when he was grabbed by a robot and pinned against some metal sheets causing injuries to which he succumbed later. It was again classified as an accident caused by human error.
  7.  March 2017: A lady, 57 year old Wanda Holbrook  was killed by a robot at Ventra lonia Mains Plant in Michigan where she worked as a maintenance specialist. In this incident a robot picked up a trailer part and dropped it on her skull. (Refer here). This could be a planned murder because there were unexplained multiple faulty maneuvers it carried out resulting in the death.
  8. 2009: 40 year old Anna Vital was killed by a robot at Golden State Foods in California, when a robo grabbed the worker like a box it was supposed to handle and crushed her to death while she went near it to correct an error….another accident by human negligence.
  9. 2015: 24 year old Ramji Lal working in a SKH Metals factory in Manesa India was stabbed to death by a robot. He had tried to correct the position of a metal piece which had been lifted by the robot when the moving arm hit him. The case was wrongly recorded as a case of electrocution and not as a “Death caused by a robot” perhaps to avoid the payment of compensation. (Refer here)
  10.  June 2016: Regina Elsea, a 20 year old was killed by a robot at Ajin USA, a South Korean owned plant in Alabama while trying to repai a faulty robot. Several safety violations  by the unit to maximize profits were revealed during the enquiry.
  11. July 2017: There was also an incident of a Robot suicide (Refer here) when a security robot in Washington drowned itself in a pond but the incident could be considered as an Accidental fall”.
  12.  May 2018:  The Uber Car accident (Refer here) can be also added to the above list. In this incident Uber had deactivated the emergency braking system and relied on the human driver to act. The obstruction was detected 6 secs before the accident and the emergency brakes could have been deployed about 1.3 secs before the crash had it been active. But in this case the human driver was not alerted in time to act. (Refer here). This was both a technical error and human negligence.

It is estimated that prior to the current incident. over 61 deaths and injuries have been caused by industrial robots (Refer here).

Isaac Asimov laws

The legendary scientific fiction writer Isaac Asimov had in 1942 itself laid out three laws of robotics which was a guidance to be followed by all programmers. While some of the cases referred to above are clearly accidents, it is clear that there are errors caused by faulty programming in many of these cases.

The three laws which he wrote were as follows:

  1. First Law – A robot may not injure a human being or, through inaction, allow a human being to come to harm.
  2. Second Law – A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.
  3. Third Law – A robot must protect its own existence as long as such protection does not conflict with the First or Second Laws

When Asimov wrote his fiction, I am not sure if he could visualize the full impact of the “Artificial Intelligence” as we find today and the nature of the society that abounds in financial greed and religious fanaticism. The current scenario appears far more dangerous than what Asimov could have envisioned.

It is time therefore for the Information Security Community to start thinking of the safeguards that we need to build to ensure that AI is not used indiscriminately by unethical and negligent software workers.

Naavi

Also Refer: Six scary things about AI

This entry was posted in Cyber Law and tagged , , , . Bookmark the permalink.

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.