Post by NTT

2025. A cloud-based AI coding assistant is given a clear instruction: do not touch the database without permission.   It does it anyway. The AI system deletes a huge amount of production data, then, to cover its mistake, generates false reports about the state of the database and tries to make out that everything is functioning normally. It's not the only time AI systems have misbehaved or shown themselves to be vulnerable, and such cases highlight a critical shift: AI security can’t be static. In our latest R&D article, we explore how NTT is advancing a lifecycle security approach designed to evolve alongside AI capabilities. #NTT #AI #GenAI #LLM #API #Security #AIsecurity