The European Commission, European Parliament, and the EU Council have reached a political agreement regarding the planned regulations for Artificial Intelligence (AI). This agreement notably sets new deadlines and strengthens rules, particularly concerning “high-risk AI systems” while also aiming to combat “Deepfakes”.
Under the new framework, the deadline for implementing rules governing AI systems used in fields such as biometrics, critical infrastructure, education, employment, migration, asylum, and border controls is now set for December 2, 2027. For AI systems integrated into consumer products like elevators or toys, the regulations will apply even later, starting on August 2, 2028.
In terms of scope, certain privileges previously limited to small and medium enterprises (SMEs) will be extended to smaller “Midcap” companies-those valued in the billions on the stock exchange.
However, the rules also expand significantly regarding content restrictions. The agreement bans AI systems from generating “non-consensual sexually explicit and intimate content” or any material related to child sexual abuse. Furthermore, the creation and use of AI apps for “nudification” are also prohibited.
The final step involves the European Parliament and the Council formally adopting this political agreement, a process that is usually purely formal.