Enhancing the performance of modern supply chains is a constant managerial challenge, particularly as performance is now defined through a multidimensional lens combining Lean efficiency and Green sustainability. While the integration of these paradigms promises synergistic benefits, it also introduces inherent contradictions. Prior work addressed these tensions by embedding the Theory of Inventive Problem Solving (TRIZ) within a Hybrid Risk Management Approach (HRMA), leveraging TRIZ to resolve conflicting Lean-Green requirements. However, classical TRIZ ideation is heavily dependent on expert reasoning, making the process time-consuming and cognitively demanding. This study extends the previous approach by introducing an AI-assisted method for contradiction resolution. It integrates natural language processing and a question-answering system to mine patents and scientific literature, automatically extracting inventive solutions to engineering contradictions. A comparative analysis is conducted using a previously studied agri-food case, assessing both classical and AI-assisted TRIZ approaches in terms of solution relevance, originality, feasibility, and time efficiency. Results highlight that the AI-augmented approach can significantly accelerate problem-solving, expand the diversity of proposed solutions, and lower the expertise barrier. On the other hand, it may also generate contradictions misaligned with the strategic goals of the problem context—underscoring the need for expert oversight. This work contributes to advancing TRIZ-based decision support by positioning AI not as a substitute, but as a facilitator that enhances and scales classical inventive problem-solving in industrial applications.