DENMARK Trends and Developments Contributed by: Simon Milthers, Thomas Bøgedal Kristiansen, Mikkel Friis Rossa and Emil Steenberg, Bech-Bruun
authors. Therefore, determining IP rights for works generated through AI involves a specific assessment based on the type of technology used and the extent of human input. This nuanced evaluation is essential to establish clear ownership and avoid potential dis- putes. Emerging technologies often come with additional regulatory considerations. By way of example, IoT devices and connected cars may fall under telecom- munications regulations subjecting the target com- pany to extensive regulatory obligations. Moreover, buyers in M&A transactions must be vigilant about data privacy risks, especially when the target com- pany employs customer profiling features or utilises new technologies such as generative AI tools or IoT devices. These technologies can raise significant data protection concerns that need to be thoroughly exam- ined during the due diligence process. Navigating AI regulations The EU is at the forefront of regulating AI and its rap- idly evolving field. The AI Regulation came into effect on 2 August 2024, but the individual rules apply pro- gressively, as follows: • 2 August 2024 – the AI Regulation came into effect (20 days after publication in the Official Journal of the EU); • 2 February 2025 – the prohibition on banned AI systems applies; • 2 May 2025 – deadline for the development of codes of practice; • 2 August 2025 – obligations for providers of gener- al-purpose AI models and provisions on sanctions apply; • 2 February 2026 – the EU issues guidelines on classification rules for high-risk systems; • 2 August 2026 – general application of the AI Regulation, including the application of high-risk obligations according to Annex III; and • 2 August 2027 – application of high-risk obligations according to Annex I (EU harmonisation legisla- tion). High-risk AI systems, such as autonomous vehicles and medical devices, will be subject to rigorous test- ing and documentation requirements. The EU AI Act
also addresses generative AI (eg, ChatGPT), mandat- ing transparency and measures to prevent the genera- tion of illegal content. The proposed EU AI Act introduces a risk-based framework to regulate AI technologies based on their potential impact on public health, safety and funda- mental rights. AI systems deemed to pose an unac- ceptable risk – for example, those involving cognitive manipulation, social scoring or real-time biometric surveillance in public spaces – will be banned. High- risk systems such as autonomous vehicles, medical devices and critical infrastructure technologies will be allowed but subject to strict regulations, including testing, data documentation and human oversight. AI systems with limited or minimal risk will face fewer requirements, mainly around transparency – for exam- ple, notifying users when interacting with AI and allow- ing them to disengage. According to the suggested legislation, companies that do not comply with the rules risk being fined. There are different levels of fines, as follows. • For violations regarding prohibited AI systems or applications, fines may amount to as much as EUR35 million or 7% of annual global turnover, depending on which amount is highest. • AI systems’ failure to meet the other requirements of the EU AI Act may result in fines of up to EUR15 million or 3% of their total worldwide annual turno- ver. • If companies provide incorrect information about the use of AI in a system or application, fines may amount to as much as EUR7.5 million or 1% of annual global turnover. The influence of AI on due diligence in M&A is poised to be profound as it becomes increasingly embedded in business operation. The due diligence process will need to evolve in order to address critical issues such as the utilisation of AI in processing personal data, the creation of IP through AI, and the potential vulnerabili- ties associated with a company’s dependence on AI systems. This evolution necessitates a more thorough assessment of AI-related risks and compliance con- siderations during the M&A process.
99 CHAMBERS.COM
Powered by FlippingBook