On 20 October 2020, the European Parliament put forward detailed legislative proposals on a civil liability regime for artificial intelligence. The European Commission will consider the proposals and reports submitted by the European Parliament, with its legislative proposal on these matters being expected in early 2021.
Changes to EU Civil Liability Framework
The proposal creates a two-fold liability regime consisting of “high-risk” AI systems and “other” AI systems. Common principles for operators of both high-risk and other AI-systems include:
- Operators cannot escape liability on the grounds that the harm was caused by an autonomous activity, device or process driven by the AI system; and
- Where multiple operators are involved, they should be jointly and severally liable, but each operator would have the right to recover part of the compensation from the other operators, in proportion to their liability, provided that the affected person was compensated in full.
For “high-risk” AI systems
- An autonomously operating AI-system is considered high-risk where it has significant potential to cause harm or damage to one or more persons in a manner that is random and goes beyond what can reasonably be expected – so we expect this to include AI such as self-driving vehicles and autonomous robots.
- Most significantly, operators of “high-risk” AI systems under this proposed Regulation would be strictly liable for any harm or damage caused by an activity, device or process driven by that AI system. Due diligence would not be a defence and any attempt to exclude liability in agreements with users would be void.
- Under the proposal, claimants will be able to recover up to 2 million euros for death or personal injury, and up to 1 million euros for economic loss or damage to property.
- Mandatory insurance would be required to be held by operators of high-risk AI technology – similar to that used for motor vehicles.
- Claims will be subject to a special limitation period of 30 years in certain circumstances.
Other AI systems
- Other AI systems would be subject to a fault-based liability regime. If the operator can establish that the AI system was activated without its knowledge (and that reasonable measures were taken to avoid this), or that due diligence was taken with regards to the AI system, the operator may be able to establish that it was not liable for the harm. Member States will be left to set the level of fines and limitation periods for these types of claims.
See the European Parliament’s resolution containing the legislative proposal here.