WIRED: How to Use AI Chatbots Without Sacrificing Your Privacy
nordicinst · x · 2026-09-22
WIRED's Andy Greenberg argues that AI chatbots have become virtual confession booths for millions, while nearly all major models collect and store that highly personal data by default—often usable for training, sharing, or handing over in lawsuits and law-enforcement requests.
Johns Hopkins professor Matt Green notes users are building "a huge profile" of themselves one question at a time, and Signal creator Moxie Marlinspike says the biggest privacy vulnerability has shifted from texting to AI interactions.
The article's advice: prefer local EU-compliant tools, review data policies, opt out of data sharing, and keep sensitive conversations offline where possible.
More from Safety
- Cisco Talos Finds CLOSEDQUORUM Malware Run by an AI Hive Mind of Four LLMs — nordicinst · 2026-09-22
- Book review: 'Silicon Sovereigns' examines AI, international law and the tech-industrial complex — ProfChesterman · 2026-09-22
- HandoffProbe ships 23 adversarial tests for A2A-to-MCP agent handoff security — HeavisideSolutions · 2026-09-22
- GovAI paper argues frontier AI evals should be embedded inside labs, not just API tests — HaydnBelfield · 2026-09-22
- NYT: Recent AI breaches at OpenAI and Google stoke concerns over increasingly capable models — nordicinst · 2026-09-22
- Report: OpenAI and Anthropic allegedly exaggerated AI security threats to protect market position — evilsocket · 2026-09-22