Experts warn that unreviewed AI-generated code in production could lead to catastrophic failures, comparing the risk to the Challenger disaster.

The New Stack is a publication covering trends and technologies in cloud-native development, DevOps, and software delivery. Developers can learn about containerization, Kubernetes, and cloud computing, as well as explore topics such as microservices architecture, serverless computing, and continuous integration/continuous delivery (CI/CD) pipelines.

The New Stack

Security experts warn that AI-generated code deployed to production without proper review could lead to catastrophic failures. The concern centers on developers using AI coding assistants to write code they don't fully understand, particularly for security-critical components. Experts recommend AI should implement validated, battle-tested libraries rather than inventing security mechanisms from scratch. Strongly typed languages like Rust offer advantages by catching errors at compile time. The risk is amplified when AI learns from existing codebases with security vulnerabilities and replicates those patterns in new code.

Vibe coding could cause catastrophic ‘explosions’ in 2026

AI should not invent security from scratch