Legit Security Reveals New Tools to Protect Software Development from Risky AI Models

With the help of new technologies, businesses can now quickly identify and fix dangerous AI models in the software development pipeline, guaranteeing the deployment of secure and compliant code.

New features have been released by Legit Security, a top platform for managing application security posture, to assist clients in identifying and reducing risks associated with unreliable AI models in software development settings. The objective of this action is to improve supply chain security for AI across the software development lifecycle (SDLC).

1. Challenges in AI Supply Chain Security:
   • Risks associated with third-party AI models in software development.
   • Potential threats like “AI-Jacking” highlighted by Legit’s research team.
2. Expanded Capabilities:
   • Identification of risks in AI models used across the SDLC.
   • Actionable remediation steps to address security vulnerabilities.
3. Empowering Security and Development Teams:
   • Flagging unsafe models with insecure storage or low reputation.
   • Coverage of market-leading AI models hub, starting with HuggingFace.
4. Complementary Features:
   • Discovering AI-generated code and enforcing policies for code review.
   • Guardrails to prevent vulnerable code from reaching production.
5. Insights from Legit’s CTO:
   • Importance of a responsible AI framework with continuous monitoring.
   • Safeguarding development practices from end-to-end against AI-related risks.

Organizations can benefit greatly from Legit Security's expanded AI detection capabilities, which include lower risk from third-party components, alerts for hazardous models, and supply chain protection for AI. Legit our dedication to offering all-inclusive solutions for securing AI in software development is demonstrated by this advancement.