Starting August 2, the EU will require labels on AI-generated content, aiming to combat the rise of deceptive media that can mislead the public. This regulation is part of the EU’s AI Act, which mandates that companies clearly identify images, audio, and text created by artificial intelligence. The significance of this move extends beyond consumer protection; it is viewed as a vital step for preserving democratic integrity in the digital landscape.
The new rules will apply to both new and existing AI systems, with existing ones granted a four-month compliance period. Companies must ensure that synthetic content, particularly that which addresses public interest, is marked as AI-generated. This is crucial as misinformation has been weaponised in political contexts, affecting public perception and trust.
While the tech industry largely supports transparency, concerns arise about the potential for overwhelming users with excessive labels. Critics argue that if everything is labelled, the effectiveness of such measures could diminish, similar to how cookie consent banners are often ignored. This could inadvertently burden businesses, particularly those less familiar with AI regulations.
As the rules take effect, companies like Google and Meta are already implementing measures to comply, including digital watermarks for identification. The broader implications of these regulations could reshape advertising and media, pushing for greater accountability in content creation and consumption, and ultimately influencing how society interacts with technology.
Source: The Guardian
