LegalWebAgent Framework

The enriched description accurately reflects the content of the arXiv paper titled 'LegalWebAgent: Empowering Access to Justice via LLM-Based Web Agents' by Jinzhe Tan and Karim Benyekhlef, published on November 28, 2025. The paper details the LegalWebAgent framework, which employs a web agent powered by multimodal large language models to enhance access to justice. It operates in three stages: the Ask Module interprets user needs through natural language processing; the Browse Module autonomously navigates webpages, interacts with page elements (including forms and calendars), and extracts information from HTML structures and webpage screenshots; the Act Module synthesizes information for users or performs direct actions like form completion and schedule booking. The framework was evaluated with a benchmark test covering 15 real-world tasks, achieving a peak success rate of 86.7%, with an average of 84.4% across all tested models, demonstrating high autonomy in complex real-world scenarios.