AI Act Leaves Stores Unreported on AI Cameras

Arnoud Engelfriet, ICT lawyer at ICTRecht, clarifies the reporting obligation for stores using artificial intelligence (AI) in cameras, specifically in the context of shoplifting. The recent AI Act does not impose a transparency obligation on stores regarding camera analyses aimed at customers. This obligation, which has been in effect since August 2, 2023, is specifically directed at situations where there is direct interaction with natural persons, which is not the case with automatic surveillance cameras.

Engelfriet explains that the AI Act is primarily designed to prevent misleading interactions, such as chatbots posing as humans. The law requires clarity regarding the use of AI systems, but the definitions are somewhat vague. For stores using AI for security reasons, there is no obligation to inform or warn customers about the presence of this technology. The interaction required for the transparency obligation is absent unless there is direct involvement from staff, for example, if a security officer interacts with a screen displaying AI analyses.

The situation is different under the General Data Protection Regulation (GDPR). Engelfriet points out that the use of AI to detect potential shoplifting falls under GDPR as automated decision-making. This raises questions about whether the regulations of Article 22 apply, which is relevant to decisions that have a significant impact on individuals. Although there remains uncertainty regarding the extent to which customers are influenced by AI decisions, there is a clear obligation for stores to provide information about the processing of personal data that leads to such detection.

According to Engelfriet, it is necessary for stores to visibly inform about the use of AI technology, among other ways by placing notifications within the store. The unique nature of this type of data processing justifies that customers are made aware of AI usage and its implications.

Read the full article from AI Insider.