| The European Parliament approved the AI Regulation on 13 March, 2024. However, a consolidated text is not yet available. It is expected to be available in the coming weeks. The AI Regulation will become effective 20 days after it’s publication in the Official Journal, while the regulations will take effect in stages. For example, provisions regarding prohibited systems will come into effect after six months, while this will only happen after three years for high-risk systems. In general, the provisions only become effective after 2 years. The regulation governs the development and use of AI regardless of the industry, making it relevant to all. The decisive factor is that the AI is used within the EU. The AI Regulation classifies AI into risk groups. Certain AI systems are completely prohibited, such as biometric categorization and social scoring. High-risk systems require a conformity assessment prior to introduction, similar to a data protection impact assessment. Public activities are currently focusing on the supervisory structure that is to exist under the AI Regulation and must be in place before the AI Regulation comes into force. In addition to national AI supervisory authorities, there will be a European coordination center. This structure is resembles the supervisory structure of the GDPR. Additionally, there will be an AI office within the Commission. We have repeatedly reported on the initiatives of the various data protection supervisory authorities, including on the topic of ChatGPT. Company users should already have a policy on the use of AI in place to regulate the processing of their personal data and the protection of any copyrights when using AI. We will get back to you as soon as the final text is available. |
|---|