LAW AND ROBOTS:
THE REALITY
Lilian EdwardsProfessor of E-Governance
University of Strathclyde
Nine WorldsAugust 2013
Robot imagery?
FICTION Terminator 2 – mp4 Robbie the Robot MariaREALITY Asimo Nao ROXXY, Geminoid F
Metropolis1927
1956/ 1965
Nao, U Herts , 2010
ASIMO (Honda)2002-
Reality robots
Industrial robots
Military robots
Robot bomb defuser
PARO
HELPER ROBOTS: Japan's population is ageing rapidly, with over 22% of the population aged 65 or older, overworked kids & few immigrant/low paid carers
MOBISERV EU project 2013
HAL exo skeleton
Robot transport – driverless cars
Sexbots
ROXXY, New Jersey, 2010
Ishiguro’s GEMINOID F2013
What this talk isn’t about Human-intelligent robots /“Strong AI” The “singularity” Hence not,
robots having “legal personality” (can sue, be accused of crimes)
Transhumanism (human mind into metal container)
Cyborgism (human mind/body enhanced by robotics) – or not very much!
What it is about Here and now robots “about as
intelligent as dishwashers”, or lobsters (Winfield)
Not necessarily or even often humanoid Features:
“intelligence”autonomous action; learning and adaptation; embodiment cf Skynet; Google;
Twitterbots/agentsmobility
Who do lawyers suddenly care about robots?
Robots now moving into consumer, domestic, and “caring” environments – not just industrial/military
Current/near current legal issues – 5-10 years away Not just about hypothetical morality, ethics or
philosophy – real problems beginning Privacy; liability; crime; evidence; road traffic law! Ethics & social issues eg under age sexbots, saving
lives with driverless cars, leaving old people alone, environment, employment impact, robowar and unequal conflict/civilian & humanitarian impact..
More interesting than Skynet!
How to regulate robots? Robots as legal category – general regulation?
Legal analogies: Person (legal personality) Slave (lesser legal personality – cf Roman law of
slavery to get round agency issues re bots) Animal? (animate, sub-legal personality,
unpredictable cf cats, some anthropomorphism) Tool – machine – car – manaufactured object –
“product liability” for consumer safety Fails to capture aspect of
learning/adaptivity/unpredictability Or the “state of the art” defense
Software?
My approach as lawyer Problems caused?
*Liability for harm caused (who is responsible?)
*Privacy (is a care robot in the home or hospice 24/7 surveillance? Control & vulnerable people?);
Criminal law (can a driverless car lose its license? Can a robot surgeon or mining robot kill? Can a robot carer “give evidence” in court re suspicious death?)
Humanitarian law (Can a robot soldier break laws of war? If a drone plane takes out civilians can US (say) be brought to ICC ?)
“General regulation”: Asimov’s Three laws of Robotics
A robot may not injure a human being or, through inaction, allow a human being to come to harm.
A robot must obey any orders given to it by human beings, except where such orders would conflict with the First Law.
A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.
3 (5) Laws for Roboticists Edwards' Three Laws for Roboticists (from EPSRC Sandpit, 2010)
1.Robots are multi-use tools. Robots should not be designed solely or primarily to kill, except in the interests of national security.
2 Humans are responsible for the actions of robots. Robots should be designed & operated as far as is practicable to comply with existing laws & fundamental rights and freedoms, including privacy.
3) Robots are products. As such they should be designed using processes which assure their safety and security (which does not exclude their having a reasonable capacity to safeguard their integrity).
.. Plus..
Robots are manufactured artefacts, so they should not be designed in a deceptive way to exploit vulnerable users (“their machine nature should be transparent”);
It should always be possible to find out who is legally responsible for a robot. (cf registered keeper of cars? Person who “signed” contract to buy robot?)
See Winfield, New Scientist, 9 May2011
Probem approach: Liability case study
A military robot kills a civilian by mistake A mining robot excavates wrong area and
landslip results damaging civilian houses A Roomba trips up an old person who hurts
herself A care robot fails to stop one child from hitting
another (?) as not in programmed remit; or report an old person swallowing too many pills (not too few)
A sex robot “learns” one kind of behaviour from person A (eg caning) that causes harm to person B
A driverless car is hacked so that it has an accident leading to economic/physical harm
Liability models - 1 Negligence
Issues : proof of fault ie breach of duty of care? Contractual exclusion of liability?
Who is liable – manufacturer – programmer – “trainer” – owner – leaser – user?)
Product liability = Strict liability for manufacturer if defect => damage. US/EU differences. In EU: state of art defense does not currently apply if the defect causing
the damage came into being after the product was put into circulation (learning)
3rd party intereference; What if eg robot car is hacked?
Liability models - 2
Animals : cf PARO Liability of custodian (cf user, or owner – not
necc the same) =strict liability or liability after some notice (1 bite) . Issues
Are robots tame or wild?? Bad analogy? Nature changes, harms v
different. Children:
VERY divergent civil/common law traditions etc. Eg Scots law , no automatic resp of parent for child’s delicts.
Robots have no ability to reach “maturity”?
Liability models - 3 Contract: Allocate liability by contract.
Fair to consumers? The new Facebook T & C? The “small print” and “shrink wrap” problems.
An insurance market. Cf cars – things we own very liekly to hurt
others or ourselves. Likely to arise for driverless cars anyway, but other robot classes?
Establishing actuarial risks v difficult. Compulsory insurance of owner?
We don’t require insurance for dishwashers – or even pets!
A few final thoughts Is there value to considering robots as a
special category for “rules” cf Asimov? Is there value to considering robots as a
special catgory even for liability? Not clear. Are we worried about the growth of domestic
surveillance by robots of our most vulnerable people? (cf Google Glass – public!)
Lack of jurisdictional harmonisation for laws will be huge issue: Japanese robot, software from US, used in EU?
Do we want THIS?