Fighting AI with AI: Leveraging Foundation Models for Assuring AI-Enabled Safety-Critical Systems
Anastasia Mavridou, Divya Gopinath, Corina S. Păsăreanu
TL;DR
The paper tackles assurance challenges in AI-enabled safety-critical domains by addressing opacity and the semantic gap between natural-language requirements and DNN behavior. It introduces REACT, which translates informal requirements into formal specifications using $LTL_f$ via FRET and $DFA$ representations, and SemaLens, which leverages Vision-Language Models to analyze, test, and monitor perception with human-understandable concepts. The authors propose an end-to-end pipeline that yields early V&V, automated test generation with traceability, and runtime semantic monitoring, thereby reducing manual effort and enabling safer autonomous systems. By aligning with standards such as DO-178C and enabling semantic reasoning and diverse test inputs, the approach aims to advance certification of learning-enabled components.
Abstract
The integration of AI components, particularly Deep Neural Networks (DNNs), into safety-critical systems such as aerospace and autonomous vehicles presents fundamental challenges for assurance. The opacity of AI systems, combined with the semantic gap between high-level requirements and low-level network representations, creates barriers to traditional verification approaches. These AI-specific challenges are amplified by longstanding issues in Requirements Engineering, including ambiguity in natural language specifications and scalability bottlenecks in formalization. We propose an approach that leverages AI itself to address these challenges through two complementary components. REACT (Requirements Engineering with AI for Consistency and Testing) employs Large Language Models (LLMs) to bridge the gap between informal natural language requirements and formal specifications, enabling early verification and validation. SemaLens (Semantic Analysis of Visual Perception using large Multi-modal models) utilizes Vision Language Models (VLMs) to reason about, test, and monitor DNN-based perception systems using human-understandable concepts. Together, these components provide a comprehensive pipeline from informal requirements to validated implementations.