The global energy crisis, driven by economic and political disruptions, has intensified efforts to transition toward a more competitive and sustainable society. This study, framed within the context of SDG 7, examines the influence of knowledge, psychological factors, and sociodemographic characteristics on two dimensions of sustainable residential energy consumption: energy efficiency and energy curtailment behavior. A quantitative survey was conducted with 1410 Romanian participants, using a structured questionnaire and convenience sampling. Descriptive and inferential statistical analyses reveal that knowledge of energy issues and the importance attributed to sustainable development goals positively influence intentions to conserve energy at home. Notably, perceived importance significantly influences the purchase of energy-efficient appliances (F = 23.01, p < 0.001) and moderately supports curtailment behaviors, as evidenced by higher adoption rates of actions such as disconnecting appliances and using natural lighting among participants with stronger pro-saving attitudes. Attitudes toward voluntary energy-saving measures also predict purchasing and curtailment behaviors, with intention playing a mediating role. Sociodemographic variables impact energy-saving behavior to varying degrees. While perceptions may differ across countries due to historical contexts, the findings provide a valuable benchmark for informing national policies and promoting voluntary energy-saving and production measures at the residential level, supporting the transition to sustainable energy.

Artificial intelligence has fundamentally altered how brands create, personalize, and deliver communication on social media, raising important questions about how these technological features shape consumer responses. This study investigates how two AI communication features (AI content optimization and AI transparency) influence brand attitude and purchase intention, with trust and perceived usefulness examined as mediating mechanisms. Drawing on the Stimulus–Organism–Response (S-O-R) framework, a process-oriented model is tested using survey data from 497 Romanian social media users and analyzed with Structural Equation Modeling. Results show that AI content optimization significantly strengthens both perceived usefulness and trust, while AI transparency contributes primarily to trust, with a more limited and context-dependent effect on brand attitude. Trust emerges as the most central mechanism, influencing perceived usefulness, brand attitude, and purchase intention, and mediating key relationships between AI features and consumer responses. Perceived usefulness and brand attitude further translate these internal evaluations into behavioral intentions. Brands should therefore prioritize relevant, well-timed AI-generated content and invest in trust through consistent communication. Theoretically, this study maps the indirect psychological pathways through which AI communication features reach consumer behavior. Empirically, it contributes evidence from an emerging European market that remains underrepresented in this literature.