Europeans could more easily demand fair compensation when robots, drones or other artificial intelligence (AI) systems cause harm, and when manufacturers fail to address cybersecurity vulnerabilities, under new EU directives adopted by the European Commission on Wednesday.

• “New technologies such as drones or AI-powered delivery services can only work when consumers feel safe and secure. Today, we offer modern liability rules that will meet this need. We will adapt our legal framework to the realities of digital transformation,” he said on Wednesday Didier Reynders, EU Commissioner for Justice.

Why EU product liability rules need to be updated

FirstThe Commission proposes to modernize the existing norms of strict liability of manufacturers for low-quality products (from smart technologies to pharmaceuticals).

Actuator Product Liability Directive (DRP)under which Europeans can claim compensation if they suffer damage caused by defective products, dates back to 1985 and no longer covers product categories that have emerged as a result of new digital technologies such as smart products and artificial intelligence (AI).

• “The current rules are not clear on how to determine who will be liable for faulty software updates, faulty machine learning algorithms or faulty digital services that are essential to the operation of the product.
• The current rules do not specify who is liable if the company significantly modifies a product already on the market or if the product was imported by the consumer directly from outside the Union.
• This makes it difficult for businesses to assess the risks of marketing innovative products and leaves victims of losses without the possibility of compensation for an increasing number of products.” – says the European Commission.

Which products will be covered by the updated rules?

According to the European Commission, the new rules will apply to everything from garden chairs to cancer drugs, from agricultural products to advanced cars, as well as software updates.

• “Just like other products, faulty software and AI systems can also cause harm, for example if they are integrated into a cleaning robot or if they are marketed as a separate digital product, such as a medical app for a smartphone. The new DRP makes it clear that affected individuals can seek compensation if software or AI systems cause harm,” explains the European executive.

The new Directive will regulate liability for artificial intelligence

SecondFor the first time, the Commission proposes concrete harmonization of national rules on artificial intelligence (AI) liability, so that victims of AI-related harm can more easily obtain compensation.

The AI ​​Liability Directive complements and modernises the EU’s civil liability framework, introducing for the first time special rules on damage caused by artificial intelligence systems.

• “Thanks to the new rules, victims of harm caused by artificial intelligence technology will be able to access compensation in the same way as for harm caused in any other circumstances.
• The Directive introduces two main measures: the so-called “presumption of causation”, which relieves victims of the obligation to explain in detail how a specific error or omission caused the harm; and access to evidence from companies or suppliers when it comes to high-risk AI.” – says the European Commission.

Why a new AI Directive is needed

Current national liability rules are not designed to handle claims for damages caused by AI-based products and services, the European Commission says.

• “For tort claims, the victim must identify the organization she wants to sue and explain in detail the fault, damages, and causation between them.
• This is not always easy to achieve, especially when artificial intelligence is involved. Systems can often be complex, opaque and autonomous, making it extremely difficult, if not impossible, for the victim to provide this burden of proof,” says the European Executive.

The new rules ensure that any type of victim (individual or business) can have a fair chance of compensation if they have suffered as a result of the fault or negligence of an AI provider, developer or user.

Next steps

The proposal of the European Commission must be adopted by the European Parliament and the European Council.

The European executive also proposes that 5 years after the entry into force of the AI ​​Liability Directive, the Commission will assess the need for strict liability rules for compensation claims related to artificial intelligence.

Photo source: Dreamstime.com