We Publish Our Third Study on the AI Act, Focusing on Article 5 & Prohibited Practices
Today, Homo Digitalis publishes its third study on Regulation 2024/1689, the now widely known AI Act, titled “Artificial Intelligence Act: Analysis of Provisions on Prohibited Practices in Article 5 of Regulation 2024/1689.”
The authors of this study are Sophia Antonopoulou, Lamprini Gyftokosta, Tania Skrapaliori, Eleftherios Chelioudakis, and Stavroula Chousou.
The aim of this Homo Digitalis analysis is to systematically approach each provision of Article 5 of the AI Regulation, related to Manipulative or deceptive techniques, Exploitation of vulnerabilities, Assessment of social behavior, Facial recognition database use, Prediction of criminal offenses, Emotion detection,Biometric categorization systems, and Remote biometric identification.
With our study, we provide targeted questions highlighting the critical aspects of individual provisions, identifying the so-called “gray areas”—points that present ambiguities, overlaps, or potential interpretative challenges. We substantiate our concerns with specific examples and pose precise questions to be addressed by the upcoming guidelines of the European Commission’s AI Office and the national legislator.
As with our first two studies (published in October and November 2024, respectively), our third study also aims to support the Ministry of Digital Governance in its mission to transpose the AI Act into Greek legislation. Additionally, through our detailed analyses and arguments, we aim to contribute to the maturation of public discourse and empower more Civil Society organizations to actively participate in it.
You can read our study, “Artificial Intelligence Act: Analysis of Provisions on Prohibited Practices in Article 5 of Regulation 2024/1689,” (available in EL) here.
Homo Digitalis participated in AI Office’s consultation on Prohibited Practices under the AI Act
In November 2024, the European Commission’s Artificial Intelligence (AI) Office launched a consultation on AI Act prohibitions and AI system definition.
The guidelines under development will help national competent authorities as well as providers and deployers in complying with the AI Act’s rules on such AI practices ahead of the application of the relevant provisions on 2 February 2025.
Homo Digitalis participated in this public consultation process by submitting our input, in an attempt to highlight challenges and provide further clarity on practical aspects and use cases.
The authors’ team of our public consultation is composed of our Director on Human Rights & Artificial Intelligence, Lamprini Gyftokosta and our members Sophia Antonopoulou and Stavrina Chousou.
You can read our input here.
Stay tuned, since our dedicated report on the AI Act and its provisions on Prohibited Practices is to be published soon!
We publish our Second AI Act Study on market surveillance authorities and the AI governance ecosystem
Today Homo Digitalis publishes its second study on Regulation 2024/1689, the now well-known AI Act, entitled “AI Act: Analysis of Provisions for AI Governance and Competent Market Surveillance Authorities“.
The writing team for the study consists of Homo Digitalis’ Director of Human Rights and AI, Lamprini Gyftokosta, and our member Niki Georgakopoulou.
The purpose of this Homo Digitalis analysis is to highlight some of the critical issues raised by the implementation of the AI governance system provisions, taking into account national structures as well as the civil society perspective.
More specifically, in this analysis we answer the following questions:
- What governance structure does the Regulation propose for AI?
- What does the concept of ‘market surveillance authority’ mean for the AI Regulation?
- What is in Regulation 2019/1020 and why should we consider its provisions together with the AI Act?
- Which Greek authorities meet the requirements set out in the two Regulations and why?
- What governance models have been adopted or are under discussion in other jurisdictions at this time?
- What are our main concerns?
- What are our main suggestions for improvement?
We recall that on 12 November, the Department for Digital Government took the first official step in implementing the AI Act by publishing the list of national authorities for the protection of fundamental rights. These principles include: The Data Protection Authority, the Ombudsman, the Communications Privacy Authority and the National Human Rights Commission.
In this regard, as early as 25 October, with our first Study “Analysis and proposals for the incorporation of the provisions on impact assessment on fundamental rights in Greece“, we had already presented detailed proposals on this issue. If you did not have time to read our Study, we invite you to see the one-page summary we prepared, specifically for the National Fundamental Rights Authorities.
The Ministry’s publication was only the first step. The next critical obligation is the institutional design of the market surveillance authorities, which must be completed by August 2, 2025, in accordance with Article 113 of the Regulation.
The second Study that we are publishing today is precisely intended to assist the Ministry of Digital Governance, which has the task of carrying out the difficult task of synthesizing this ecosystem in Greece, but also, with our detailed analyses and arguments, to help mature the public debate and enable more civil society organizations to actively participate in it.