While consumer protection of the environment has become a concern in recent years in all parts of the world, an attitude-behavior gap remains that hinders turning pro-environmental attitudes into sustainable purchases. While AI, specifically personalized AI, is slowly becoming a digital marketer's go-to, the ability of AI, and specifically of generative Artificial Intelligence (AI) and Large Language Models (LLMs) to leverage this innovation in digital marketing, and beyond conversion and engagement metrics, has yet to be fully explored in developing-economies settings. This paper suggests a conceptual model that will combine both models to gain a better understanding on how to effectively use AI-driven personalisation in order to help shift climate-conscious marketing from awareness raising to actionable. The framework also includes protections from algorithmic "greenwashing" and privacy issues, making ethical transparency a condition to maintaining consumer trust. It builds and discusses four propositions on the relation between AI personalization design choices and strategies for reducing the attitude–behaviour gap, and provides implications for scalable and technology-driven methods of climate communication for marketers, policymakers and environmental education advocates.