The European Union is on the brink of passing groundbreaking legislation on artificial intelligence (AI), potentially the first of its kind globally. This legislation, known as the EU AI Act, could introduce regulations on a wide range of AI applications, from homemade chemical weapons to copyright theft of music, art, and literature. The Act, which is currently under negotiation between MEPs, EU member states, and the European Commission, is expected to be adopted by parliament and become law by early next year.
One of the key areas of debate is the use of AI-powered live facial recognition. While member states argue that this technology is crucial for border security and public order, MEPs believe that real-time facial recognition cameras in public spaces infringe on privacy rights. The Act also aims to address the potential unknown threats posed by AI, such as the creation of biohazards. Developers or owners of AI tools could be held accountable and face fines of up to 6% of their revenue or be banned from the EU if their AI model is capable of producing something illegal.
The Act also includes provisions to protect the creative sector. AI companies will be required to submit lists of data sources to the European Commission regularly, which will help deter unauthorized use of data and creative content. This measure is expected to benefit musicians, scientific researchers, and authors by providing them with legal protections against plagiarism.
Lastly, the Act will require tech companies to regularly publish data on their electricity consumption. This provision is in response to reports about the significant energy consumption of AI training runs. The aim is to increase transparency and shape public policy regarding energy use in the AI sector.