Before robots replace us entirely, we will first have to learn to live and work with them… And that means robot insurance.
Do you remember the SNL skit from 1995 with Sam Waterston as a “compensated endorser” for Old Glory Insurance? Old ladies in a living room are chatting, disappointed at how the world is falling apart with all the gangs and rap music… And robots.
One of the ladies says, “They’re everywhere!” Another chimes in, “Darren and I have a policy with Old Glory Insurance that covers us in case we’re attacked by robots.”
Will marauding robots from the SNL skit become reality like the Simpsons predicting a Trump presidency? Who knows…!? But as they encroach on new territories beyond controlled, industrial environments, robots will present many new challenges for insurers.
Robot Safety Concerns
While industrial robots have been around since the 1950s, they have been confined to cages and cordoned off areas to protect co-workers. However, advancements in robot technologies, combined with lower costs, are making close quarter collaboration between robots and humans more commonplace. Robots are working in warehouses, performing surgery, making pizza, driving and helping to enforce the law. But as “robocowos” or “cobots” (robot co-workers) do more and different types of work, it’s safe to assume that new forms of robot insurance will become a common type of property and liability coverage.
New robot insurance forms are being created by pioneering insurance carriers, such as Lexington, a division of AIG, to address risks related to artificial intelligence (AI), advances in machines, such as self-driving cars, autonomous robots, drones and other technologies.
However advances in artificial intelligence and robotics are happening faster than any one carrier can handle. Governments, legislators and international standards bodies like ISO1 are struggling with the issues of how to deal with emerging technologies. Meanwhile, the robotics industry is expected to grow to $151 billion by 20202, and risks will become even more complex. As such, the potential for mistakes when “human meets robot” will increase exponentially.
Some scientists predict an intelligence explosion, much like the Big Bang. In this scenario, scientists expect that artificial intelligence systems will become recursive, allowing AI and robotic machine learning systems to build new intelligent robots themselves, constantly improve and fix themselves if necessary, with no need for human involvement.
Scientists do not agree when this may happen, with estimates ranging from within 30 years to never. However, if a global cycle of robot and artificial intelligence is set in motion, will we be able to stop it, or know the cause, or who to blame, for a robot related claim?
What is Robot Insurance?
Robots already do work for us, but are are expected to do more in industries from transportation to the law. As puny humans creating gods, how can we protect ourselves from the predictable risks and reduce our exposure to claims from artificial intelligence and robot problems?
Robot insurance is made to protect businesses and individuals from claims related to mistakes, malfunctions and design flaws from robot-related accidents. Robot insurance coverage types are currently limited, with only a few carriers offering coverage And much like cyber insurance coverage, policy coverage ranges from scant to extensive. As with cyber, there is not enough actuarial data to properly price the risk of many robot applications and their exposures. Robot insurance coverage includes aspects of three common insurance types:
- Cyber Insurance: Cyber insurance risks, such as malware, ransomware and hacking are important endorsements to include in robot insurance because almost anything that has an Internet connection is exposed to outside interference.
- Technology Errors and Omissions Insurance: Robots are driven by software. When software fails to perform as expected it may be covered by an errors and omissions policy. Technology errors and omissions (E&O, also known as professional liability) insurance covers risks related to robotic products and services.
- Product Liability Insurance: Product liability insurance protects against liability claims related to artificial intelligence architectures and robotic products for individuals, businesses and governments.
As robots become more common and embedded in our daily lives, what types of claims could arise?
What are Robot Risks?
Companies are developing new applications for robots on a daily basis. Applications range from robots that can make pizza or have sex or play ping pong, to consumer focused robots, such as the Roomba automated vacuum robot made by iRobot, to business, industrial, security and military applications. Prototypes of more advanced, humanoid robots are being developed by Honda and Boston Dynamics that may be used in complex manufacturing and military applications that are currently only the domain of humans.
The potential bad situations that could arise from the aforementioned applications are left only to the imagination.
While robots are often more predictable than humans at some tasks, robots are also known to malfunction due to software or hardware errors, as well as from human error. When a robot fails to work as it should the result can be property damage, injury or even death. In such an event, robot insurance would be there to provide coverage for bodily injury to others, property damage, medical bills, business interruption, and loss of income.
In 2015 a contractor at a Volkswagen plant in Germany was killed when a stationary robot grabbed and crushed him against a metal plate.3 A security robot drowned itself in a fountain, and a Russian robot escaped from its enclosure and caused a traffic jam when it went out onto the nearest street.
Robot Ethics… Who is Responsible?
Ethicists have long pondered questions of right and wrong in certain situations. But when it comes to robots, or software, how will they decide when the situation results in someone dying?
For instance, imagine an autonomous vehicle with a 60 year old passenger driving along when a child runs out into the road after his ball. There is not enough road for the car to stop… So the software must make a split second decision whether to continue forward, or swerve to the right or left. If the car goes forward, it will hit and kill the child. If it swerves to the right, it will hit a telephone pole and certainly kill its 60 year old passenger. If it swerves to the left, it will hit oncoming traffic and kill passengers of an unknown number.
Someone is going to die… But what is the ethical thing to do? And who or what will be responsible for indemnification in the aftermath? The vehicle manufacturer? The software maker? The driverless vehicle’s 60 year old passenger?
The convergence of technology, ethics and responsibility puts insurance companies in uncharted territory.
The insurance industry and legal experts expect that as robots and artificial intelligence become more commonplace, the liability will shift away from consumers and be absorbed by the automotive industry and to manufacturers of robotics products.
- ISO/TS 15066:2016 https://www.iso.org/standard/62996.html
- Tractica, 19 Nov. 2015: https://www.tractica.com/newsroom/press-releases/global-robotics-industry-to-surpass-151-billion-by-2020/
- A German news agency reported that prosecutors were unclear whether they should bring charges in the case, and if so, against whom?