MEPs voting on rules for interact between humans and artificial intelligence and robots

Happening today

MEPs are to vote on the first comprehensive set of rules for how humans will interact with artificial intelligence and robots.

They are voting on the EU Draft Report with recommendations to the Commission on Civil Law Rules on Robotics by the Committee on Legal Affairs


Key elements of the Draft Report (with para numbers) – and my comments

1.       Adopting Asimov’s 3 laws of Robotics (L) – arguably a weak idea, because they were drafted for dramatic purpose, not legislative

2.      EU planning to pay role in establishing basic ethical principles in development, programming, use of robots and AI (N) – possible since US and Asia have no similar initiative

3.      Start with civil liability, and consider if strict liability (i.e. no fault, insurance-based) is better (P) – leaves it to injured parties to sue, instead of imposing criminal sanctions

4.      Autonomous robots may have liability for injury / harm (T, U) – massive questions in tort law, because negligence liability requires the defendant to have taken ‘reasonable care’, and what is reasonable for a robot?

5.      Autonomous robots may be able to enter contracts (X) – enforcement of contract requires sanctions, what sanctions would be relevant to a robot?

6.      General definition of smart robot includes physical support – this should not be a requirement because it fails to account for very smart and powerful robots that have no physical form, which are working on financial markets, information and communications, etc

7.      Robots may be able to own intellectual property in works they create –

8.      Standardisation, safety and security (13) – this is useful and necessary

9.      Harmonisation of laws on autonomous vehicles (15) – this is useful and necessary

10.   Care and medical robots (16 – 17) – does it excuse doctors from liability?

11.   Human enhancement (18) – if an implant has autonomy, what rights should it have vis a vis the human?

12.   EU framework on drones (19) – this is already in progress

13.   Employment forecast and suggestion of universal basic income (20 – 23) – necessary to start this discussion

14.   Liability (24 – 31) with compensation to be provided by compulsory insurance – helpful because the robot will not be able to compensate an injured human; designers and makers may also welcome this as protection for themselves