I tested Google’s upcoming Gemini Nano 4 — its faster, smarter AI isn’t what I expected

Google’s upcoming Gemini Nano 4 claims to offer faster on-device AI for smartphones, but preliminary testing reveals potential security and privacy risks in its local execution model. Early adopters and enterprise users deploying AI-driven applications on mobile devices are likely to be affected by vulnerabilities in data handling and inference processes. Immediate mitigation is advised to prevent unauthorized access or data leakage.