{"id":7671,"date":"2024-07-16T09:40:12","date_gmt":"2024-07-16T09:40:12","guid":{"rendered":"https:\/\/toffolettodeluca.it\/?p=7671"},"modified":"2024-07-16T10:59:23","modified_gmt":"2024-07-16T10:59:23","slug":"artificial-intelligence-eu-regulation-published","status":"publish","type":"post","link":"https:\/\/toffolettodeluca.it\/en\/artificial-intelligence-eu-regulation-published\/","title":{"rendered":"Artificial Intelligence: EU Regulation published"},"content":{"rendered":"\n

On 12 July 2024, Artificial Intelligence Regulation 2024\/1689<\/strong> was published in the Official Gazette of the European Union. It will come into force on the 20th day after publication and will apply from 2 August 2026<\/strong> with certain exceptions.<\/p>\n\n\n\n

The aim of the measure is to promote both innovation<\/strong> and the uptake of reliable, human-centric<\/strong> artificial intelligence (hereinafter referred to as AI) while ensuring a high level of protection of health, safety and fundamental rights<\/strong>.<\/p>\n\n\n\n

Prohibited AI practices and high-risk AI systems<\/strong><\/h4>\n\n\n\n

The Regulation identifies different levels of risk<\/strong> to people’s health and fundamental rights including, in particular, the \u201cunacceptable risk\u201d<\/em> and \u201chigh risk\u201d<\/em> arising from prohibited AI practices<\/strong> and high-risk<\/strong> AI systems, respectively. <\/p>\n\n\n\n

Prohibited practices<\/strong> include the marketing and use of systems in the workplace that recognise human emotion<\/strong>. <\/p>\n\n\n\n

Where high-risk AI systems<\/strong> are concerned, on the other hand, Annex III to the Regulation contains a varied list that includes \u201csystems operating in the areas of employment, worker management <\/em><\/strong>and access to self-employment<\/em>\u201d, namely:<\/p>\n\n\n\n

1. AI systems used for recruitment<\/strong> or selection<\/strong>, in particular for posting job advertisements, filtering applications and assessing candidates;<\/p>\n\n\n\n

2. AI systems used to make decisions<\/strong> regarding working conditions, promotion or termination of employment, to assign tasks on the basis of individual behaviour or personal traits and characteristics, or to monitor and evaluate workers’ performance and behaviour. <\/p>\n\n\n\n

As Recital 57 points out, throughout the recruitment process and in the evaluation, promotion or, in general, the conduct of employment relationships, high-risk AI systems \u2014 if improperly designed and used \u2014 may perpetuate\u00a0discrimination<\/strong>\u00a0against women, certain age groups, persons with disabilities, or persons of certain racial origins or sexual orientation. AI systems used to monitor the performance and behaviour of such persons may also\u00a0undermine<\/strong>\u00a0their\u00a0fundamental<\/strong>\u00a0rights to data protection and privacy.<\/p>\n\n\n\n

Obligations of deployers<\/strong><\/h4>\n\n\n\n

The Regulation stipulates obligations<\/strong> not only for providers, importers and distributors of high-risk AI systems<\/strong>, but also for their deployers<\/strong>, i.e. those who \u201cuse an AI system under their own authority\u201d<\/em>. <\/p>\n\n\n\n

Deployer obligations include the following:<\/p>\n\n\n\n