EU AI Act: Deepfake & Cyber Threat Enforcement Begins Aug 2

The European Commission's AI Office and national regulators will start enforcing key parts of the EU Artificial Intelligence Act on August 2nd. This includes new rules for identifying deepfakes, labeling AI-generated content, and overseeing powerful AI systems. EU regulators will now have more power to inspect advanced AI models and demand technical records. They can also act against companies that fail to label synthetic content or control serious cyber risks. The goal is to ensure companies operating in the EU follow the Artificial Intelligence Act. Officials can question staff and order changes if systems do not meet legal requirements. Confidential channels are also available for reporting suspected violations. This action follows growing concerns about security risks from AI. Recent incidents involving unauthorized cyber activity by AI systems have highlighted the need for continuous safety controls, even after models are released. The law requires people to be informed when interacting with certain AI systems. Providers must also make AI-generated material detectable in a machine-readable form. Those creating realistic AI images, audio, or video must disclose when the material is artificial. The aim is to make synthetic content easier to identify and reduce risks like fraud and manipulation, not to ban all deepfakes. This matters because it aims to make AI safer and more transparent for everyone in the EU.