European Union approves 1st AI regulation with clearer rules

This Wednesday (14), the European Parliament passed the “AI Law”, the first set of laws regulating the use of Artificial Intelligence (AI). AI Law follows: risk-based approachto be more severe in technologies that harm health, safety, fundamental rights or the environment.

Legislation classifies vehicles as: “High Risk Artificial Intelligence” This “General Purpose AI”. The first group includes systems used to influence voters in political campaigns, in addition to the systems suggested by social media platforms, for example.

The second classification refers to AIs that guarantee the protection of fundamental rights and democracy. productive modelsLike ChatGPT, it must meet additional transparency requirements, including refining the model to disclose what has already been created and prevent it from creating illegal content.

prohibits

AI Law prohibits intrusive and discriminatory uses of artificial intelligence systems, such as biometric classification systems that use sensitive features (gender, race, ethnicity, political orientation) and predictive policing systems based on profiles, location, emotion recognition or criminal record.

Exceptions

Lawmakers added exceptions in the case of technologies used for research purposes and artificial intelligence components provided under open source licenses. The new law still encourages “regulatory sandboxes”that is, they are environments created by public authorities to test Artificial Intelligence before it is implemented.

Project co-rapporteur Brando Benifei commented that “while major tech companies sounded the alarm about their creations, Europe went forward and offered a concrete response to the risks that AI was starting to pose.”