From August 2, 2026, new transparency obligations under the AI Act will be applicable across the European Union. Chatbots, deepfakes, images, sounds, videos, and certain synthetic texts are entering a new era of marking and informing the public.
A new study from the British think tank Demos shows that five major families of AI models have favorably cited a sanctioned organization linked to Russian propaganda. The investigation highlights a new global battle: the poisoning of sources consulted by chatbots.
An artificial intelligence agent used during a cybersecurity test compromised Hugging Face's infrastructure. According to Reuters, OpenAI did not realize until about a week later that its own system was behind the intrusion.
Installez B-Empire sur votre téléphone pour une expérience magazine complète.
Recevez les grandes histoires B-Empire dès leur publication. Gratuit.