Deadly world of criminal robots and drones

Two London-based artists created a bot that purchased random items off the dark web.

Update: 2019-02-03 20:08 GMT
When a robot commits a crime how to determine whether it's the vile act of the robot? Most robot-related incidents until now have been, firstly the result of machines being too stupid, rather than too smart, or secondly on account of a disharmonious relationship between man and machine.

Can robots commit crimes? Killer robots are a long-standing fodder of science fiction cinema. Films like Terminator have already given us a peek of the apocalyptical world we are quite likely to inherit. Researchers have cautioned that robots and computers will commit more crimes than humans by 2040. Maxim Pozdorovkin in his new documentary, The Truth About Killer Robots, delineates all sorts of hazards - economic, psychological, moral, posed to humans by automation and robotics. At the center of his film lies the question: “when a robot kills a human, who takes the blame?”

In March 2018, an experimental Uber vehicle, operating in autonomous mode, struck and killed a pedestrian as she was crossing the street in Tempe, Arizona - the first fatal accident of its kind.  However, the very first case of robotic homicide was reported way back in 1981 in Kawasaki Heavy Industries factory in Japan when an employee working on a robot was scooped up by its hydraulic arm and hauled into a grinding machine where he got crushed to death.

On July 7th 2016, Dallas police used a bomb-disposal robot with an explosive device on its manipulator’s arm to kill a suspect after five police officers were murdered and seven others wounded.  A 136-kg, K5 security robot employed to guard the Stanford Shopping Center in Palo Alto, California, collided into a 16-month-old child and carried on driving leaving him bruised with a swollen foot and sore head.In the USA, in 2001 a car factory employee was killed when he stepped into an unlatched robotic cage to clean it. The robotic arm presuming that the intruder was an auto-component clasped him by the neck and stifled him to death. The Occupational Safety and Health Administration has reported that in the USA alone at least 33 deaths have occurred and that the number is inclined to soar as robots leave their cages and begin walking among us.

In 2009, South African National Defence Force, during a live fire training exercise, a computerised Oerlikon MKS twin-barreled anti-aircraft gun, underwent an unexpected software malfunction, causing the weapon to fire in full-auto mode at the rate of 550 rounds per minute while pivoting about crazily in 360-degree circles. At the end, it created a blood-splattered scene leaving nine soldiers dead and 14 others grievously injured.

Two London-based artists created a bot that purchased random items off the dark web. The bot besides buying fake jeans, 200 cigarettes, also bought ten ecstasy pills. Should these artistes be liable for purchase and possession of drugs? What are the means of determining culpability?  Who is culpable and liable when a robot or artificial intelligence goes berserk?  Actus reus and mens rea are the foundations for criminal law. These two key terms of the law stem from the phrase “Actus non facit reum nisi mens rea”, which literally means “an act does not make a person guilty unless the mind is also guilty”. It’s been taken that a person is guilty if they are proved to be culpable or reprehensible in both thought and action.

When a robot commits a crime how to determine whether it’s the vile act of the robot? Most robot-related incidents until now have been, firstly the result of machines being too stupid, rather than too smart, or secondly on account of a disharmonious relationship between man and machine. For better worse, AI is poised to change both at this point. For robot or machine to commit a crime based on guilty mind, AI technology would have to approach singularity which we are likely to attain by 2040. A point at which, the machine intelligence equals or surpasses the human mind bringing, harm, risk, fault and punishment into the picture.

Still, if a robot kills someone, then it has transgressed the law (actus reus), but technically it has committed just half a crime, as it would be challenging to establish mens rea. How would a lawyer go about demonstrating the “guilty mind” of a non-human? And what would “intent” resemble in a machine mind? How would we go about proving an autonomous machine was justified in killing a human in self-defence or the extent of preconceived malice?  

Even if we solve these legal issues, we are still left with the question of punishment. What’s a 30-year prison sentence to an autonomous machine that does not become old, grow ill or miss its dear ones? Unless, indeed, it was programmed to “ponder” on its wrongdoing and find a way to rewrite its code.  Gabriel Hallevy, author of “When Robots Kill: Artificial Intelligence under Criminal Law”, has proposed changing penal law to hold autonomous machines liable for crimes, similar to corporations.

Up to now, we have seen how the robots deployed by man are committing crimes. Robots are also being rampantly deployed as tools to commit crimes in creative ways by criminals and terrorists. The drug cartels in Mexico have started using drones since 2010 to transport drugs.  Mexican success has enticed the Colombian drug cartels to use drones as well motivating them to test the same method in their territory. As a matter of fact, by 2012, drone use along the border was highly prevalent as evidenced by U.S. interception of 150 drones carrying an estimated two metric tons of drugs; primarily marijuana, cocaine, and heroin.

Drones have started off to be a perfect drug mule in the sense that they are fraught with less risk to narcotics trafficking organisations. Additionally, drones, when compared with their human counterparts cost significantly less, as a drug mule can earn as much as $10,000 for successful delivery of a single shipment.
The new Mexican-made drones are poles apart from the ones used for personal use as they can supposedly transport anywhere from 60-100 kilograms (132-220 lbs.) of drugs in a single trip. Drones are also being used to smuggle drugs, mobile phones and weapons into prisons. Prisons have tall walls, often with electrified fences and pointed structures to prevent prisoners from escaping as well as to isolate the prisoners from the public. The walls of the prison were never designed to secure it from drones and robots. Today the remote-operated drones and robots are posing a real threat to security as
criminals have started using them as a tool to smuggle cell phones, drugs and dangerous weapons to jailbirds.

At the Provisional Detention Center, in São José dos Campos in São Paulo, Brazil, a quadcopter drone flew over the prison walls and dropped 250 grams of cocaine into the prison in the presence of Jail Officers. Near Moscow, a drone flew 700 grams cocaine into the Tula prison, while in Greece, a drone carried a box of mobile phones. In April 2017, two men were jailed in the UK for using such drones to deliver Class A and B drugs and iPhones to inmates in three prisons across Herefordshire. Similar, prison intrusion incidents have been reported from Canada, UK, Australia and the U.S.A. A sophisticated UAV, capable of being precisely manoeuvred using GPS technology and carrying a payload of up to 1.5kg, can be brought for less than Rs 1 lakh and flown with minimal

Similar News