Artificial Intelligence (AI) is becoming increasingly integrated into decision-making processes, yet it remains unclear how individuals perceive and evaluate the fairness of AI systems in different resource allocation scenarios. This study examined how variations in distributive fairness (decision outcomes: favourable vs. unfavourable; within) and procedural fairness (transparency level: low vs. balanced vs. high; between) influence perceived fairness, trust, acceptance, and intentions towards AI-based allocation decisions. Using an experimental vignette design, participants (N = 929) evaluated four AI-based resource allocation scenarios (e.g., kindergarden spots). In line with prior research, our linear mixed-model results showed that outcome favorability was the strongest predictor for all dependent variables. Surprisingly, in contrast to our expectations and prior studies, the degree of transparency, operationalized as the amount of information shared with participants, did not significantly affect perceived fairness, trust, acceptance, or usage intentions. These findings highlight the robustness of outcome favorability effects across different resource allocation contexts while simultaneously raising questions about the appropriate operationalization of transparency in such settings. In this regards, we conclude that future research may benefit from differentiating between information quantity, information content (e.g., process vs. algorithm), and complexity, to better capture the multifaceted role of transparency in shaping human responses to AI.