Supreme Court (Ukraine): Artificial Intelligence Responses Are Not Recognised as a Source of Reliable, Acientifically Proven Information
Artificial intelligence can be a useful and helpful informative tool in the field of justice, but it cannot replace either a judge or the principles of relevance, admissibility and reliability of evidence provided for in Chapter 5 ‘Evidence and Proof’ of the Civil Procedure Code of Ukraine.
This conclusion was reached by a panel of judges of the Commercial Court of Cassation within the Supreme Court.
The city council filed a lawsuit against a limited liability company to amend the land lease agreement and oblige it to recalculate the rent.
The lawsuit was motivated by the need to bring the terms of the lease agreement into line with the city council’s decision.
By a decision of the commercial court, which was left unchanged by the decision of the commercial court of appeal, the claim was partially satisfied, namely, amendments were made to the land lease agreement.
The Supreme Court upheld the decisions of the lower courts.
The Supreme Court rejected the defendant’s argument that the court of appeal had unreasonably rejected the request to examine electronic evidence, namely the responses of two artificial intelligence systems – Grok (developed by xAI) and ChatGPT (developed by OpenAI) – confirming the literal interpretation of subparagraph 6 of paragraph 23 of the agreement, and noted the following.
The panel of judges emphasised that technology should only be used to support and strengthen the rule of law. Technology can only be used to support and assist courts and judges in the proper administration and determination of proceedings. Decisions must, explicitly or implicitly, be made solely by judges. This cannot be delegated or performed by technology; judicial autonomy must be respected.
In the case in question, however, the participant uses artificial intelligence technology not as a means of facilitating the administration of justice, but rather to challenge (question, appeal) conclusions already made by the court.
Artificial intelligence responses are not recognised as a source of reliable, scientifically proven information, so the court’s rejection of the request to examine them as evidence in the case is lawful.
Post Comment