The adoption of AI-enabled systems and services depends on user trust. To achieve this, the design and commercialization of AI must prioritize transparency, fairness, and accuracy to create a positive user impression. This research sheds light on how developer-specific characteristics can impact users' perceptions of AI systems, particularly when they have no prior experience with them. Using a controlled experimental design, this research demonstrates that users expect AI provided by non-profit organizations to be more transparent and fairer but not necessarily more accurate than AI provided by for-profit organizations. The results are explained by perceptions of organizational greed and impressions of warmth. Implications are discussed for algorithm aversion and the managerial practices of AI providers.