Apiiro launches Guardian Agent to secure AI-generated code

Application security posture management company Apiiro Ltd. has announced the launch of Guardian Agent, a new artificial intelligence application security agent designed to guard AI models and prevent vulnerable or noncompliant code before it is generated. The offering aims to help enterprises scale AI-native development while reducing risk and improving compliance, according to a recent report by Silicon Angle.

Guardian Agent addresses the challenge of rapidly expanding application attack surfaces, which have grown up to 10 times due to AI's impact on software development. Apiiro argues that traditional application security scanners are insufficient for current realities, being fragmented, noisy, and prone to false positives. The new agent works by continuously understanding software architecture, attack surface, runtime exposure, and organizational policies. It then uses this context to guard AI assistants and development workflows in real time, rewriting prompts to prevent the generation of insecure code.

Key features include adapting secure prompts as requirements evolve, eliminating manual threat modeling, and translating complex security decisions for various stakeholders. By focusing on understanding and prevention rather than reactive measures, Apiiro aims to combat alert fatigue and improve overall security posture.

Source: Silicon Angle