This study developed and evaluate an artificial intelligence (AI) counselling chatbot to mitigate maladjusted behaviours among secondary school students in Niger State, Nigeria. Guided by a Research and Development design integrated with a within-subjects (repeated measures) experimental approach, the population of the study 98,018 senior secondary school students, sample size comprised of 360 students while the sampling techniques utilize multistage sampling technique. The chatbot addressed aggressiveness, phobia, inferiority complex and anxiety. Needs analysis revealed high emotional challenges, limited access to counselling and strong receptivity to chatbot interventions. ADDIE model guided systematic design, development, implementation and evaluation and validated for content, technical and field usability. Reliability was established with Cronbach’s alpha (0.771). Pre/post assessments analysed with Wilcoxon Signed-Rank and Mann-Whitney U tests showed significant improvements in aggressiveness, phobia, inferiority complex and anxiety. Findings revealed strong need for chatbot support, with over 45% reporting emotional struggles and 50% willing to try chatbot counselling. Effectiveness was evident: aggressiveness reduced (≥88% agreement), phobia managed (≥86%), inferiority complex alleviated (≥87.5%) and anxiety controlled (≥90%). Gender differences were minimal, with males reporting slightly higher benefits. Hypothetically the findings revealed that counselling chatbot significantly reduced aggressiveness (M = 44.48; 25.44), phobia (M = 25.34; 44.33), inferiority (M = 25.30; 44.96) and anxiety (M = 25.03; 44.03), all p < .001. Gender differences were non-significant: aggressiveness (U = 15,885.50, p = .750), phobia (U = 16,118.50, p = .934), inferiority (U = 16,077.50, p = .901), anxiety (U = 15,676.50, p = .595). This implies that the counselling chatbot was highly effective in mitigating maladjusted behaviours among secondary school students in Niger State. The intervention significantly reduced aggressiveness, phobia, inferiority complex and anxiety. Importantly, the results further indicate that the counselling chatbot was gender-neutral, exerting equivalent influence on both male and female students.