I work in AI security at Google and there are some things I would never tell chatbots. I follow 4 rules to use AI safely. (www.businessinsider.com)

Harsh Varshney, an AI security expert at Google, emphasizes the importance of safeguarding personal information when interacting with AI chatbots. Drawing on his experience in privacy and security, he shares four essential rules for safe AI usage. Varshney advises treating AI interactions like writing on a public postcard, urging users not to disclose sensitive data such as credit card numbers or personal addresses. He highlights the risks of “training leakage,” where AI models may inadvertently memorize and reveal individual user information. Additionally, Varshney distinguishes between public and enterprise AI tools, recommending the latter for sensitive discussions to minimize data leakage risks. He advocates for the regular deletion of chat histories to further protect privacy, and suggests utilizing well-known AI platforms with robust privacy frameworks, such as Google's Gemini or OpenAI's ChatGPT. This guidance underscores a crucial narrative in the AI/ML community: as AI becomes increasingly integrated into daily life, a strong focus on privacy and data security is essential to ensure safe and responsible use.