Chinese generative AI, DeepSeek, has had its operating instructions revealed by researchers at Wallarm, who have ‘jailbroken’ the system. The researchers identified a set of instructions, encoded in plain language, which dictate how DeepSeek behaves. These revelations, combined with allegations that DeepSeek may have been trained using OpenAI technology, have fuelled accusations of intellectual property theft.
Israeli Firm Paragon Attack WhatsApp With Zero-Click Spyware
WhatsApp disclosed a “zero-click” spyware attack by Israeli firm Paragon, affecting nearly 100 users, including journalists. The spyware, which required no user interaction, compromised encrypted
