AI Liability Directive: Council of the European Union identifies potential changes

The Council of the European Union's Commission Services has provided considerations for potential substantive amendments to the proposed Artificial Intelligence Liability Directive (AILD) (see here). The AILD is a proposed and complementary piece of legislation to the EU AI Act, providing a liability framework so that "persons harmed by artificial intelligence systems enjoy the same level of protection as persons harmed  by other technologies." (see our AILD summary here). 

In particular, the AILD is seen as complementary to the EU AI Act, by providing persons harmed by AI systems greater rights, in this case, to claim compensation.  Further, the AILD contains server links with the AI Act and, as a result of the AI Act being updated since the AILD was drafted, updates are now being considered for the AILD's drafting.  

The considerations for substantive amendment to the AILD - which do not constitute a Commission position - can be summarised as:

1. Damages covered under the AILD
   1. In short: Are new obligations in the AI (since previous drafts which the AILD was based on) intended to protect against the damage a potential claimant may suffer and whether breach of those obligations would result in a ‘Rebuttable presumption of a causal link in the case of fault’ under the AILD (Art 4). 
   2. Some of those new obligations in the AI Act which raise questions about whether and to what extent they are covered by AILD include:
      1. AI Literacy - the AI Act requires “providers and deployers of AI systems to take appropriate measures to ensure a sufficient level of AI literacy of their staff and other persons dealing with the operation and the use of AI systems on their behalf” (AI Act Art. 4);
      2. investigations into causes of risks caused by AI systems (AI Act Art. 20);
      3. obligations for providers of general purpose AI models (GPAI)  (AI Act Art. 53);
      4. obligations for providers of general purpose AI models (GPAI)  with systemic risk (AI Act Arts. 51 and 63)
2. Disclosure of evidence and rebuttable presumption of non-compliance
   1. In short: Whether the right under the AI Act for an individual to an explanation of decision-making means changes are required to the AILD provisions regarding disclosure.
   2. AILD would require that Members States ensure that national courts are empowered to order disclosure of 'relevant evidence' provided the claimant i) presents facts and evidence sufficient to support the ‘plausibility of a claim for damages’, and ii) first takes all proportionate attempts at gathering the relevant evidence from the defendant. 
   3. Where the Defendant fails to meet such an order ‘a national court shall presume the defendant’s non-compliance with a relevant duty of care, … that the evidence requested was intended to prove for the purposes of the relevant claim for damages’, although the Defendant can rebut that presumption.
   4. However, how should the right an individual has to an explanation under the AI Act - and any evidence produced - interact with these provisions in the AILD?

If you would like to discuss how current or future regulations impact what you do with AI, please contact Tom Whittaker, Brian Wong, Lucy Pegler, David Varney, or Martin Cook. 

For the latest on AI law and regulation, see our blog and sign-up to our AI newsletter.