This weekly update covers major data privacy and security developments focusing on agentic AI risks and protections. A popular Chrome extension called Urban VPN Proxy was caught secretly collecting everything people typed into AI chatbots like ChatGPT and Claude, stealing conversations from over 7.3 million users. Meanwhile, experts warn that autonomous AI tools are creating new security problems in offices, as employees accidentally share sensitive company information through AI assistants. The United States government is also pushing new rules to help keep AI systems safer. New York State passed the RAISE Act, which requires companies developing the most powerful AI models to report safety problems to state officials within 72 hours. This protects against the growing threat of AI being used to write phishing emails and malware faster than ever before. Additionally, the FBI warned that criminals are using AI-generated voice messages to trick people into giving away passwords and personal information. In the United Kingdom, Apple disabled a security feature after government pressure to weaken phone encryption. These developments show that as AI becomes more powerful and independent, protecting people's privacy and security requires new rules, better company practices, and stronger defenses.