This paper explores how sustainability-themed advertising influences consumer perceptions and purchase intentions in the fashion retail sector, specifically focusing on the Italian market. Using a mixed-method approach, the study integrates quantitative survey data, biometric facial recognition, and natural language processing (NLP) to assess the impact of advertising elements such as credibility, transparency, and emotional engagement. A sample of 120 Italian consumers viewed a sustainability-oriented promotional video and responded to structured questions and open-ended prompts while being recorded via webcam. Statistical analyses, including regression and machine learning models (Lasso, Random Forest), revealed that perceived credibility (β = 0.41, p < 0.001) and transparency (β = 0.31, p < 0.01) significantly predict purchase intention, moderated by environmental concern. Biometric data showed that 68% of participants expressed joy during segments featuring quantifiable sustainability claims. Sentiment analysis using BERT confirmed that 77% of open-text responses were positive, often citing “authenticity,” “clarity,” and “impact.” Demographic differences emerged, with younger consumers showing greater responsiveness to emotional appeals while older participants prioritized factual content. The findings offer important managerial implications for tailoring sustainability communication, suggesting that brands must combine evidence-based messaging with emotionally resonant storytelling. The study concludes that credibility and emotion function synergistically in sustainability advertising and recommends using AI-based tools for campaign design and testing. Limitations include the single-video format and short-term measurement, with future research encouraged to explore longitudinal effects, cross-cultural dynamics, and multi-format media strategies.