After several years of negotiations and countless debates, the final version of the EU AI Law is finally due to be adopted in mid-July 2024. It will then take another 24 months before far-reaching obligations apply to most companies. The AI Law will be the first regulation of algorithms based on “machine learning”. However, authorities around the world have already issued numerous recommendations and guidance documents on the use of AI. We have created an overview of some of the most important guidelines.
AI regulatory framework before the AI Act
Artificial Intelligence (AI) is penetrating every sphere of life and raising complex issues ranging from huge opportunities to technical risks and ethical considerations. There are many legal challenges due to the processing of personal data, use of copyrighted material and trade secrets. There are also concerns that companies with large market shares will stifle innovation by investing heavily in AI start-ups.
The economic, social and individual impacts have given rise to a comprehensive regulatory environment for AI systems. In this regard, the authorities have recently issued several guidelines and recommendations on data protection, IT security, competition law, etc., which relate to both the technological and economic development framework and the use of AI systems.
Summary of government statements on AI
We have grouped these statements here by area. Essentially, they fall into the following areas:
Data Protection: Data protection authorities address the interaction of data protection requirements with AI technologies, which often include transparency of data processed by AI systems, data minimization, and other core principles such as GDPR.
Cybersecurity: In this regard, the guidelines primarily focus on how to protect AI systems from threats, in particular including guidelines for identifying security risks and mitigation strategies against attacks on AI applications in critical infrastructure.
Competition Law: Competition regulators will investigate the extent to which AI systems or their use can foster innovation and influence market practices that fall under antitrust laws. The evaluation of such collaborations demonstrates efforts to ensure fair competition in the context of modern technologies.
These guidelines are intended to ensure that AI systems are ethical, trustworthy, and developed and deployed in accordance with legal requirements. Authorities will continue to base their decisions on the existing guidelines even after the AI Act is passed.
The outlook for AI law
The AI Law will have far-reaching effects on both the development and use of AI. However, its actual impact on providers and users is still unclear in many areas. The first publications by authorities such as the “NIST AI RMF Generated AI Profiles” and the French Data Protection Supervisory Authority (CNIL) guidelines on the use of AI systems show first signs. However, the AI Law will be further substantiated in the coming days by voluntary codes of conduct, guidelines and delegated acts. In particular, the AI Law provides for:
- Voluntary Code of Conduct: These will be driven by the AI Office and EU Member States with the aim of independently examining the objectives and requirements of the AI Law. As the AI Law contains a large number of obligations, it remains to be seen what their practical relevance will be.
- Committee Guidelines: These will be even more relevant for the practical implementation of the obligations in practice: they will provide valuable information on the scope of prohibited systems, the exact obligations of providers of high-risk systems, the practical implementation of transparency obligations, etc. To date, there is still a lot of legal uncertainty in these areas.
- General AI model practice guidelines: These have been developed by the AI Office for specific obligations and are intended to assist providers in preparing risk assessments and transparency reports.
- Transparency guidelines: These are issued by the AI Office and facilitate the identification and labelling of artificially generated or manipulated content.
More about AI law
We offer a wide range of content covering the latest developments in AI. We cover important topics such as data protection, copyright, employment law, IT security and competition law and provide practical guidelines for companies and developers. See an overview here. The next episodes of the international AI webinar “Tech Me Up!” will take place on July 9 and September 19. In the next episode we will address issues of governance of AI law. After the summer break we will have a data protection authority as guest. You can register for the webinar here.
