EU to tighten accountability for AI-caused harm

The European Commission has proposed new rules to ease how individuals and companies file claims against manufacturers of AI-based products. Reuters reports.

Under the proposal, victims would be able to seek compensation for harm to health, privacy and property caused by the fault of AI technology producers or suppliers. The rules would also impose liability for discrimination in the hiring process conducted by algorithms.

\”We want the same level of protection for victims of damage caused by AI as for victims of ordinary technologies,\” said Justice Commissioner Didier Reynders.

The new rules would ease the burden of proof for victims through a ‘presumption of causality’. Victims would only need to show that non-compliance by the producer or user of certain requirements caused harm. They could then link these facts to the AI technology in their suit.

Under the ‘right of access to evidence’, victims can obtain disclosure of information about high-risk AI systems. This, the authors of the law say, should help establish who is liable for the damage.

The Commission also announced updates to the Product Liability Directive. Under the amendments, manufacturers face penalties for all unsafe products, including software and digital services.

Users can seek compensation when software updates undermine the security of smart-home devices. Those with unsafe products outside the EU can file for compensation with the manufacturer’s representative in the EU.

Before the bill becomes law, it must be approved by the European Parliament.

Earlier in July, the EU adopted laws, regulating tech giants in privacy, algorithms and competition.

In June the EU updated the rules for tackling misinformation and deepfakes online.

Subscribe to ForkLog news on Telegram: ForkLog AI — all the news from the world of AI!