[DigitalToday reporter Chi-gyu Hwang (황치규)] Information security company Winstechnet said on Thursday it launched SNIPER AIVAX, an AI prompt security solution, to respond pre-emptively to security threats stemming from the spread of generative artificial intelligence.
The company said SNIPER AIVAX automatically detects sensitive data, including personal information, corporate secrets and technical information, contained in prompts entered when using generative AI services. It also blocks the data before it is sent to external AI services.
It said this defends against unintentional information leaks, as well as malicious prompt injection attacks and attempts to steal data.
The company said it applies a three-layer security system covering prompt security, Model Context Protocol security and network security. It said the system provides detailed analysis and policy-based controls at the prompt stage, protects communications between AI models and links with existing network security to implement multi-layered defence.