The study explores the potential of AI enhanced demand prediction for recycled and upcycled products, aiming to bridge the gap between sustainable production and market adoption. We propose a study utilizing machine learning algorithms to analyze historical sales data, market trends, and consumer sentiment, thereby forecasting demand for eco-friendly products with unprecedented accuracy. Leveraging business analytics and social media influence data, we propose a novel framework to: (1) compare consumer preferences and purchase patterns for recycled and upcycled products against their traditional counterpart (2) identify key factors driving demand for sustainable products, (3) compare demand patterns between ARIMA and LSTM models, and (4) analyze the impact of social media influence on consumer preferences. By predicting future demand with enhanced accuracy, this framework aims to empower businesses to optimize production, pricing, and marketing strategies, ultimately enabling a more sustainable and circular economy. The study holds significant implications for both businesses and consumers seeking to minimize environmental impact while catering to evolving market trends.