Many people find it hard to read complex online texts, especially those with ADHD, dyslexia, memory difficulties, or who are learning a new language. This study looks at how AI can help make digital content easier to read. A prototype was developed that uses Puppeteer to extract website content and LLaMA 3.2 to simplify the text based on Easy-to-Read (E2R) guidelines. An online survey with 54 general readers and a usability test with nine participants who have reading challenges were conducted. Results showed that most people preferred the simplified version. In the online survey, 69% chose the AI version, and it was read about 33 seconds faster. In the usability test, reading time dropped by over 82%, and all participants answered the questions correctly. Many found the simplified version easier to read and understand, though some felt it missed details or sounded too casual. The findings suggest AI can support accessibility by reducing cognitive load, but the tone still matters. The study highlights both the benefits and limitations of using AI for real-time text simplification on websites.