The Member States and the European Parliament agree on the basis on which the future regulation of Artificial Intelligence (AI) will be developed, establishing red lines and obligations for services such as Chat GPT.
The European Union has approved a pioneering regulation on artificial intelligence (AI) that seeks to set a precedent in the regulation of this area worldwide, similar to what happened at the time with the General Data Protection Regulation. This regulation comes after more than two years of deliberations and seeks to regulate technological progress to guarantee security and the preservation of citizens’ fundamental rights, for which purpose it establishes a risk-based classification of AI systems.
In this risk-based system, it prohibits uses of AI with “unacceptable risk” that threaten fundamental rights, such as voice assistants that may negatively influence minors or the recognition of emotions in the workplace or school. The indiscriminate collection of facial images is also banned.
Likewise, a distinction is made between high-risk applications in areas such as health, justice, or education, which will be subject to different security and accuracy guarantees, and other systems with limited risk, which will also be subject to lesser transparency obligations.
The regulation will also create a governance structure through the creation of institutions and agencies at EU and state level, in charge of oversight and enforcement of a sanctioning regime that will also be defined in the regulation.
In conclusion, it is expected that in the coming weeks the technical details will be finalised before the text is submitted for formal approval by the institutions, the final draft of which we could see in 2024.