Artificial Intelligence

The use of artificial intelligence (AI) has been a major trend in many sectors for some years now, as companies leverage the many opportunities for optimising processes and boosting productivity. AI systems can be deployed in almost every conceivable field, including by banks and media firms, in recruiting, in the sport and healthcare sector, in the automotive industry and in the energy sector. Alongside the seemingly unlimited opportunities for digital transformation, however, automation of business models and AI-supported decision-making also carry legal risks that companies need to address.

28/02/2024

Launch of the European Legal AI "Noxtua"

AI startup Xayn and international business law firm CMS develop Europe's first sovereign Legal Large Language Model and Legal Copilot Lawyers can use Noxtua to analyze, review, and summarize legal texts  Noxtua specializes in legal texts and meets the high requirements of legal professional secrecy and data protection reg­u­la­tion­sNox­tua is a sovereign and secure European alternative to US AI tools   

Press releases

11/03/2024

Looking ahead to the EU AI Act

Introduction The European Union is preparing for the imminent adoption of the world’s most significant legislation on Artificial Intelligence, solidifying its position as a pioneer among global legislators. This initiative aims to establish and reinforce the EU’s role as a premier hub for AI while ensuring that AI development remains focused on human-centered and trustworthy principles. To expedite the achievement of these goals, on 8 December 2023, after three days of debate, the European Parliament and the Council of the European Union finally reached a provisional agreement on the “Proposal for a Regulation laying down harmonised rules on artificial intelligence” (the so-called AI Act), which aims to ensure that AI systems placed on the European market are safe and respect the fundamental rights and values of the EU. Subsequent to this provisional agreement, technical refinement of the AI Act continued to finalise the regulation’s details and text. The final vote of the European Parliament on the AI Act will take place at 13 March 2024. Since the European Parliament's Committees on the Internal Market and Consumer Protection (IMCO) and on Civil Liberties, Justice and Home Affairs (LIBE) have endorsed overwhelmingly the proposed text, the approval of the European Parliament can be expected. After a long and complex journey that began in 2021 with the European Commission’s proposal of a draft AI Act, this new regulation is expected to be passed into law in spring 2024, once it has been approved by the European Parliament and the Council of the European Union . The AI Act aims to ensure that the marketing and use of AI systems and their outputs in the EU are consistent with fundamental rights under EU law, such as privacy, democracy, the rule of law and environmental sustainability. Adopting a dual approach, it outright prohibits AI systems deemed to pose unacceptable risks while imposing regulatory obligations on other AI systems and their outputs. The new regulation, which also aims to strike a fair balance between innovation and the protection of individuals, not only makes Europe a world leader in the regulation of this new technology, but also endeavours to create a legal framework that users of AI technologies will be able to comply with in order to make the most of this significant development opportunity. In this article we provide a first overview of the key points contained in the text of the AI Act1This article (including the relevant citations below) is based on the latest draft available on the Council’s website. The AI Act remains subject to possible further refinement, but not as regards content, and the text referred to for this article should be considered as the closest to the one that will be voted on by the EU Parliament. footnote that companies should be aware of in order to prepare for the implementing regulation.

Publication

Sustainable AI through data security and data quality

From a simple chatbot to a self-learning system, the benefits of AI to a company should not be detrimental to social interaction. Like with humans, AI must therefore avoid discriminatory decisions that automatically disadvantage certain groups of people, for example. Discriminatory results are often due to an AI system being trained using an inaccurate dataset. Such results can be avoided by incorporating sustainability aspects in corporate strategy, applying human checks and maintaining a high standard of data security and data quality. A sustainable strategy for dealing with AI minimises the risk of legal problems when using it.

Legal regulation of AI

Policymakers and legislators have long been aware of the potential benefits of AI for the common good, as well as the risks posed by unregulated or uncontrolled AI. Examples include the Digital Strategy of the European Union (EU) and the German government's coalition agreement. In addition to the proposal for an international AI convention, the Artificial Intelligence Act (AI Act) will be an important pillar of the Digital Strategy at the EU level. This legislation provides for transparency obligations in the case of some AI systems, e.g. chatbots and deepfakes. In accordance with the AI Act, further strict requirements apply, graded according to the risk posed by the particular AI system to the rights and interests of individuals and of society as a whole. These obligations include the requirements that the AI system has to meet, risk management, data governance (including in relation to training), monitoring, cyber security and transparency. They may also involve bans on companies. Even companies that only use, rather than create, systems will be subject to special requirements to ensure monitoring and provide information about incidents. In addition to the AI Act, the EU Commission has also proposed an AI Liability Directive. This provides for compensation to cover damage or losses caused by AI, aimed at promoting AI systems that minimise harm.

Complex systems raise complex legal issues

Due to the complexity and future importance of the topic, companies would be well advised to address the numerous legal issues that can result from the use of AI as they relate to their own innovative business model so that they avoid legal problems:

• What statutory regulations apply to AI now and in the future? 
• Who is liable for AI errors and damage caused by AI? 
• How can discrimination by AI be avoided and do automated decisions have to be reviewed by humans? 
• What should be taken into account when using AI for algorithmic management?
• What employment law aspects should be considered when using AI in a corporate setting?
• Is there a legal obligation to state that AI is being used?
• What should be taken into account with regard to the dataset with which the AI is trained? 
• What are the cyber security and data protection requirements and, in the case of machine learning, what are the requirements with regard to the quality of data? What data protection requirements apply to the processing of personal data by AI systems?
• What should be taken into account when using AI in the Internet of Things (IoT), in Industry 4.0 and in the Metaverse? 
• How does the use of AI affect intellectual property? Who is the inventor or originator of an innovation created using AI? How should prompts be judged according to law?
• What measures can be taken to assure the protection of trade secrets when AI systems are used? 

How can AI systems operating in the public cloud be prevented from utilising questions that include trade secrets when generating answers for other users (third parties)?
The experts on our interdisciplinary teams have in-depth knowledge both of the technology and of the associated legal issues and they support you around designing and implementing your business models in a legally compliant manner.

You can find publications of our lawyers on this topic here (in German): 

• Braegelmann / Kaulartz | Rechtshandbuch Smart Contracts | 1. Auflage | 2019 | beck-shop.de
• Bullinger / Vonthien, KI-Training' und Urheberrecht, Mitteilungen der deutschen Patentanwälte 2023, Seite 446
• Gerecke, Social Media und Recht: Einige urheberrechtliche Gedanken zu generativen KI-Modellen, GRUR-Prax 2023, Seite 381
• Grützmacher, Der Entwurf der EU-Verordnung für Künstliche Intelligenz, in: Buck-Heeb / Oppermann, Automatisierte Systeme, 1. Auflage, 2022
• Grützmacher, Verantwortung und Haftung bei (digitalen) autonomen KI-Systemen | AnwBl Online 2022, 498-505
• Grützmacher, Autonome Systeme und KI – vom vollautomatisierten zum autonomen Vertragsschluss?
• Grützmacher, Die zivilrechtliche Haftung für KI nach dem Entwurf der geplanten KI-VO — Potentielle zivilrechtliche Auswirkungen des geplanten KI-Sicherheitsrechts: ein neues Schutzgesetz i.S.v. § 823 Abs. 2 BGB am Horizont
• Kaulartz / Braegelmann | Rechtshandbuch Artificial Intelligence und Machine Learning | 1. Auflage | 2020 | beck-shop.de
• Kaulartz / Merkle | Rechtshandbuch Künstliche Intelligenz | 2. Auflage | 2024 | beck-shop.de
• Knappertsbusch, Owlit – Quo vadis, KI? KI-Entwürfe auf EU-Ebene und deren Bedeutung für Arbeitgeber
• Knappertsbusch / Gondlach | Arbeitswelt und KI 2030 | 1. Auflage | 2022 | beck-shop.de
• Knappertsbusch / Jares, ChatGPT und arbeitsrechtliche Aspekte – Expertenforum Arbeitsrecht (#EFAR) (efarbeitsrecht.net)
• Knappertsbusch / Vonthien a.o. | Generative KI in Unternehmen | 1. Auflage | 2024 | bitkom e.V.