Austria: Artificial Intelligence Act of the European Union

 

Artificial intelligence (AI) is the technology behind systems such as ChatGPT or Gemini AI. Since such systems were made available to the general public more than a year ago, legal questions have repeatedly arisen regarding the possible areas of application of these systems and the further legal implications. In most cases, the existing legal basis is insufficient to answer these questions.

In April 2021, the European Commission presented a draft regulation laying down harmonised rules on artificial intelligence (Artificial Intelligence Regulation), thus initiating the ordinary legislative procedure at European Union level. On 09.12.2023, the European Parliament has now agreed with the Council on a draft regulation to ensure that AI systems placed on the market and used are safe and that existing fundamental rights and the values of the Union are safeguarded.

The draft regulation is limited to the absolute minimum requirements, in the Commission’s view, necessary to address the risks and problems associated with AI. In particular, the aim is to prevent technological development from being unduly restricted or otherwise hindered.

On the basis of the above-mentioned considerations, the Council and the European Parliament have agreed to define areas of application in which the use of AI systems will be prohibited. For example, biometric categorisation systems that use sensitive characteristics (e.g. political, religious or philosophical beliefs, sexual orientation, race), the untargeted reading of facial images from the Internet or from video surveillance systems to create facial recognition databases will be prohibited in the future, but also, for example, emotion recognition in the workplace and in educational institutions. However, there are special regulations for criminal prosecution.

The use of artificial intelligence also plays an important role in the financial sector, for example, and AI systems used to assess credit scores or the creditworthiness of natural persons are classified as high-risk AI systems and these systems must therefore meet minimum standards. AI systems that select potential applicants for open positions are also classified as high-risk AI systems.

The minimum standards for high-risk AI systems require risk mitigation systems, high quality datasets, activity logging, detailed documentation, clear user information, human oversight, and a high level of robustness, accuracy, and cybersecurity.

Compliance with the Regulation is to be ensured through fines for breaches of the AI Regulation, with the amount of fines being set as a percentage of the global annual turnover of the company concerned in the previous financial year or a predetermined amount, whichever is greater.

The agreed text of the future regulation will now have to be formally adopted by the European Parliament and the Council, which will then be published as a regulation of the Parliament and of the Council and will enter into force 20 days after publication in the Official Journal of the European Union. Two years after its entry into force, the Regulation will then be applicable.

The European Union’s Artificial Intelligence Act is the world’s first comprehensive legal framework for artificial intelligence and will pose challenges for legal practitioners due to its novelty alone.

 

For further information, please contact:

Luca Trondl, Lawyer

Zumtobel + Kronberger, Salzburg

e: office@eulaw.at

t: +43 662 62 45 00

 

#WLNadvocate #Austria #law #legal #lawfirm #lawyer #network #article #corporatelaw #AI #artificialintelligence #ChatGPT #EUlaw #EU #EuropeanUnion#GeminiAI #AIregulation #technologylaw