This correlational research study explored how ICU nurses’ reliance on artificial intelligence–based clinical decision support systems (AI-CDSS) relates to decision regret, and whether trust in AI moderates that relationship. A total of 250 nurses (response rate 63%) from three Egyptian hospitals completed validated instruments: the Healthcare Systems Usability Scale (HSUS), measuring AI-CDSS reliance; the Decision Regret Scale (DRS), scoring regret from 0 (low) to 100 (very high); and the Trust in AI Scale. Nurses reported moderately perceived AI reliance (mean score of 78.6), moderate regret (mean 38.5), and moderate trust (mean 13.9). Correlation analysis showed three key findings: (1) higher AI reliance was associated with greater trust (r = 0.51, p < 0.01); (2) higher reliance was linked to lower regret (r = −0.42, p < 0.01); and (3) higher trust was also linked to lower regret (r = −0.33, p < 0.01). Together, reliance and trust explained 27% of the variance in regret scores across participants. Trust significantly moderated the relationship, further reducing regret when reliance was high. These findings echo two decision-making theories. Janis and Mann’s framework suggests trusted aids reduce internal conflict, while Lazarus and Folkman’s stress appraisal model highlights that tools seen as reliable lower emotional strain. At the bedside, AI-CDSS may help reduce nurses’ decision-related regret, but benefits depend on fostering trust through transparency, training, and alignment with professional values.