AutoSpec: An Agentic Framework for Automatically Drafting Patent Specification

2025-09-23

AI Summary

Patent drafting incurs high costs and lengthy timelines, further hindered by stringent data confidentiality requirements, domain-specific technical complexity, and long-context dependencies—key bottlenecks for automation. To address these challenges, we propose a locally deployable multi-agent patent drafting framework built upon open-source large language models. It integrates task decomposition, retrieval-augmented generation, and tool-augmented reasoning to modularize and coordinate the intricate drafting workflow. The entire system operates offline, ensuring zero exposure of sensitive patent information. Evaluated under a rigorous, patent-attorney-designed protocol—including both automated metrics and expert human assessment—our framework significantly outperforms existing baselines in technical accuracy, legal compliance, and textual conformity, demonstrating strong practical viability for real-world deployment.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Generation
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Patents play a critical role in driving technological innovation by granting inventors exclusive rights to their inventions. However the process of drafting a patent application is often expensive and time-consuming, making it a prime candidate for automation. Despite recent advancements in language models, several challenges hinder the development of robust automated patent drafting systems. First, the information within a patent application is highly confidential, which often prevents the use of closed-source LLMs for automating this task. Second, the process of drafting a patent application is difficult for even the most advanced language models due to their long context, technical writing style, and specialized domain knowledge. To address these challenges, we introduce AutoSpec, a secure, agentic framework for Automatically drafting patent Specification. Our approach decomposes the drafting process into a sequence of manageable subtasks, each solvable by smaller, open-source language models enhanced with custom tools tailored for drafting patent specification. To assess our system, we design a novel evaluation protocol in collaboration with experienced patent attorneys. Our automatic and expert evaluations show that AutoSpec outperforms existing baselines on a patent drafting task.

Problem

• Automating patent drafting to reduce time and costs
• Addressing confidentiality concerns with open-source language models
• Overcoming technical challenges in long, specialized patent writing

Innovation

• Agentic framework decomposing drafting into subtasks
• Open-source language models enhanced with custom tools
• Secure automated patent specification drafting system