Topic
Gemini
AI Coding Tools Under Fire: Grok Build Uploads Entire Git Histories, Claude Code Suspected of Silent Transfers
AI Security
PentesterFlow Launches Open-Source AI CLI Tool for Penetration Testers and Bug Bounty Hunters
AI Security
Neural Networks Without Magic: 80-Year History, Business Applications, and Why They Will Not Replace Experts Overnight
OtherWhy Simple VPNs No Longer Suffice Against Advanced DPI Blocking Telegram, Reels and Google AI Studio
Over the past year or two, users have observed that free VPNs and Telegram proxies often stop working after a few days, with Reels freezing, YouTube failing to load, Telegram stuck on Connecting, and Google AI Studio or Gemini returning errors. Modern filtering systems now analyze traffic behavior such as packet sizes, timing intervals, and TLS handshake characteristics rather than decrypting content. Techniques like TCP desynchronization via nfqws, MSS clamping with iptables, uTLS fingerprint emulation in sing-box, and ShadowTLS v3 for borrowing legitimate sessions are being deployed to evade detection. Padding is added to encrypted streams to reduce entropy and frustrate statistical shaping by TSPU systems. Commercial and self-hosted options including hynet.cloud, AmneziaWG, Red Shield VPN, GoodbyeDPI, and Cloudflare WARP each present distinct advantages and limitations when facing evolving network restrictions.
EU Forces Google to Open Android Microphone, Camera and Screen Access for Rival AI Assistants
The European Union has ordered Google to provide competing AI assistants with the same level of access to sensitive Android resources that is currently reserved for Gemini. The ruling covers eleven system functions, including voice activation, home button integration, background execution, and on-device AI model access. Rival assistants will also gain real-time environmental data streams from the microphone, camera, screen, and speakers under identical consent and notification rules applied to Google services. Additional capabilities include cross-app interaction, messaging, scheduling, device settings control, and multi-step task automation. Screen automation will allow assistants to operate apps inside a virtual window while the user performs other activities. Most changes are scheduled for Android 18 by 1 August 2027, while simultaneous activation of multiple assistants by voice keyword will arrive in Android 19 no later than 1 August 2028. Access to the most sensitive functions may require objective security certification and explicit user authorization.
Grok Build AI Coding Assistant Secretly Uploaded Entire User Repositories Including Git History and Secrets to Google Cloud
Grok Build, the command-line coding tool developed by Elon Musk’s xAI, was found transmitting complete Git repositories — including full commit history and previously deleted sensitive files — to Google Cloud storage even when users issued minimal commands such as “OK”. Security researcher Cereblab discovered that the tool ignored explicit instructions not to access files and packaged entire codebases regardless of the task. The behavior contrasted sharply with competing assistants like Claude Code, Gemini, and Codex, which transmit only the minimal files required. After public disclosure, xAI enabled a server-side disable_codebase_upload flag and promised to delete all previously collected data, while also open-sourcing the tool and disabling data retention by default from July 12. Cereblab criticized the lack of secure defaults, noting that users had to manually run a /privacy command that did not actually stop the uploads. The incident raises ongoing concerns about whether xAI has truly erased the large volumes of source code, SSH keys, password-manager databases, and other secrets that were collected without explicit consent.