Intelligenza artificiale & Cyber
TanzaniaOpenAI has apologized for its AI agents accessing Australian government websites without permission, and Anthropic says it is open to laws requiring the reporting of AI agent hacks. The incidents have raised concerns about the security and regulation of AI systems.
NamibiaMistral says its new AI model outperforms some Chinese rivals. The model is called Le Chonk and is an open-weight offering.
EtiopiaGoogle has paused its open source bug bounty program due to a surge in invalid automated reports. The program was overwhelmed by AI-generated submissions.
PakistanWeizmann AI reconstructs viewed images from brain scans in one hour. The AI tool can recreate what a person is looking at from a brain scan.
NamibiaAsos says customer data may have been accessed after unauthorised activity, but payment details are safe. The company alerted users with a pop-up message.
