EU to mandate AI content labeling
What's the story
The European Union (EU) will start enforcing its new artificial intelligence (AI) transparency regulations from Sunday (August 2). The rules require companies to clearly label content created by AI, including deepfakes. The main objective is to help users easily identify whether the online material they encounter is real or fake. Non-compliance with these labeling requirements could lead to hefty fines for companies.
Gradual enforcement
New rules to be implemented in phases
The EU's comprehensive AI law will be implemented in phases.
Starting Sunday, companies will have to ensure their AI systems, like chatbots, clearly identify themselves as such.
If an image or text is created using AI, it should be explicitly labeled as such. This could involve adding watermarks and other indicators for easy identification of AI-generated content.
Disinformation concerns
Generative AI can create disinformation at unprecedented scale
An EU official said generative AI can create disinformation at an unprecedented scale, tailored for specific audiences, and spread it quickly.
They added that the technology is "making it increasingly difficult for all of us to distinguish what is real from what is synthetic."
The new rules are aimed at preserving citizens' ability to trust what they see, hear, and read online.
Content regulation
Deepfakes a major concern for the EU
The EU is particularly concerned about deepfakes, text, images, videos, and sounds that look real but are created or manipulated by AI.
The new rules apply to content created for professional purposes. However, individuals using AI for personal purposes will not be affected.
Texts meant to inform the public on general interest issues will also have to be labeled if created using AI without human editorial oversight.
Industry response
Tech companies already implementing their own labeling systems
In anticipation of the new regulations, major tech companies have already started implementing their own labeling systems.
TikTok, for instance, has been required content creators to label AI-generated images, audio, and video for years.
The platform claims that over three billion content items have been labeled through its tools and detection technology.
Similarly, Meta has introduced an "AI Info" label on Instagram and Facebook posts using the technology.
Regulatory complexities
Google's warning about potential pitfalls of new regulations
Google has signed the EU code of conduct on AI transparency and is working with other companies like NVIDIA, OpenAI, and Apple to develop digital tagging tools.
However, Google's Karen Massin has warned about "regulatory complexity" that could be counterproductive.
She said "If online content is flooded with overlapping AI labels and legal disclosures, it becomes harder for people to get the clear context they need."