Creating Piping and Instrumentation Diagrams (P&IDs) is crucial for complex engineering systems, but the manual process is time-consuming and prone to errors. Imagine a future where engineers simply describe a system in plain English, and an AI generates the corresponding P&ID automatically. New research introduces an "AI copilot" that takes precisely this agentic approach, converting natural language descriptions into accurate P&ID diagrams. This innovative copilot uses a multi-step process involving "plan" and "execute" AI agents. First, the AI plans the steps required to create the diagram based on the text description. Then, it executes these steps, adding equipment, instruments, and connections while maintaining context from previous actions. This approach contrasts sharply with simpler AI methods that struggle with the complexity and structured nature of P&IDs. The copilot's strength lies in its ability to generate a textual representation of the diagram, adhering to the DEXPI standard, a crucial format for interoperability in engineering software. This textual representation is then translated into a visual diagram, like those created in Microsoft Visio, which engineers can further refine. The research shows this method significantly outperforms standard AI approaches in both accuracy and completeness of the generated diagrams. The copilot isn't just a one-shot tool. It's designed for iterative design, allowing engineers to build complex systems subsystem by subsystem, refining the design with each iteration. Furthermore, the generated diagram and its associated data unlock downstream tasks like automated analysis and report generation, paving the way for a more streamlined and efficient engineering workflow. While promising, the system currently relies on carefully structured prompts, a limitation researchers aim to address with future prompt automation. The scarcity of real-world P&ID data also poses a challenge for broader testing and validation. Despite these challenges, this research signifies a leap towards a future where AI significantly empowers engineers, automating tedious tasks and unlocking new possibilities in complex system design.