Commission proposes landmark new rules on liability for AI systems

EU

On 28 September 2022, the European Commission proposed the draft AI Liability Directive (AILD Draft) on adapting non-contractual civil liability rules to artificial intelligence as a potential solution for the problem of civil liability for AI systems in the EU. Because civil liability in the EU is regulated at the member state level, if the same artificial intelligence system is used in more than one country, different substantive and procedural laws apply to the assessment of potential liability for damages. Currently, complying with different rules can therefore mean additional legal tasks and significant cost increases for providers and users of AI systems providing cross-border services. If passed, the proposed AILD will create benchmark regulations across the EU on civil liability and AI.

Scope and applicability

The AILD addresses cases where AI systems do not comply with the requirements of the draft AI Act due to the fault of the providers or the AI systems users.

The AILD lays down common rules for the following:

  • the disclosure of evidence on high-risk AI systems to enable a claimant to substantiate a non-contractual fault-based civil law claim for damages; and
  • alleviating the burden of proof in the case of non-contractual fault-based civil law claims brought before national courts for damages caused by any AI system whether it is high risk or non-high risk.

Disclosure of documents evidencing compliance

Upon the request of the claimant or potential claimant, a court may order the disclosure of relevant documentary evidence about specific high-risk AI systems that are suspected of having caused damage.

The prerequisites for such an order by the court are:

  • the potential claimant must present facts and evidence sufficient to support the plausibility of a claim for damages in support of the request; and
  • the claimant must have undertaken all proportionate attempts at gathering the relevant evidence from the defendant.

Where the defendant fails to comply with an order in a damages claim to disclose or to preserve evidence, the court will presume that the defendant has also failed to comply with the duty of care in relation to the operation of the relevant AI system – since compliance with the duty of care may be demonstrated if evidence is not provided or preserved.

If the AILD is adopted, this will mean in practice that providers of high-risk AI systems will be required to prepare and keep under review, throughout the operation of the system, documentation relating to the AI system in question so that this material can be used in any court proceedings.

Presumption of causal link in case of fault

A claimant seeking damages may be confronted with the challenge of establishing whether the damage caused by the AI System is attributable to the fault of the AI System provider or user. This fault is typically a breach of the obligations contained in the draft AI Act.

According to the presumption introduced by the AILD Draft, national courts will presume, for the purposes of applying liability rules to a claim for damages, the causal link between the fault of the defendant and the output produced by the AI system or the failure of the AI system to produce an output, in the following situations:

  • the claimant has demonstrated or the court has presumed the fault of the defendant, or the fault of a person for whose behaviour the defendant is responsible, in non-compliance with the duty of care laid down in EU or national law (e.g. sectoral legislation like MDR, IVDR regulations) that is directly intended to protect against the damage that occurred;
  • it is reasonably likely, based on the circumstances of the case, that the fault has influenced the output produced by the AI system or the failure of the AI system to produce an output;
  • the claimant has demonstrated that the output produced by the AI system or the failure of the AI system to produce an output gave rise to the damage.

If the claim for damages concerns an AI system that is not high risk, the AILD presumption will only apply where the court considers it excessively difficult for the claimant to prove the causal link.

For more information on the AILD proposal, contact your CMS client partner or local CMS experts.

Article co-authored by Anna Zsófia Horváth