Evaluating LLMs for Career Guidance: Comparative Analysis of Computing Competency Recommendations Across Ten African Countries

LLMs show significant gaps in understanding local African contexts: When asked about computing career requirements across ten African countries, the six LLMs tested demonstrated only 35.4% contextual awareness on average, often failing to recognize country-specific factors like local technology ecosystems, language requirements, and national policies—a critical limitation for students seeking relevant career guidance.

Ethics and responsible AI competencies are inconsistently recommended: While all models consistently covered technical skills like cloud computing and programming, there were "notable differences" in how they addressed non-technical competencies, particularly ethics and responsible AI use—suggesting that LLMs may not be adequately preparing African computing graduates for the ethical dimensions of their work.

Open-source models outperformed proprietary ones in contextual awareness: Llama 3 (4.47/5) and DeepSeek (4.25/5) demonstrated stronger contextual awareness and better balance between technical and professional skills compared to proprietary models like ChatGPT-4 (3.90/5) and Claude (3.46/5), indicating that development approach may influence how well LLMs serve diverse global contexts.

Being open-source doesn't guarantee responsible AI outcomes: Mistral AI, despite being open-source, scored 0.00/4 on contextual performance, demonstrating that "development philosophy alone does not guarantee contextual responsiveness"—a reminder that responsible AI requires more than just open-sourcing models.

Current LLM career guidance risks perpetuating digital colonialism in Africa: The study draws on Digital Colonialism Theory to highlight how LLMs trained primarily on Western data may reinforce existing inequalities rather than serve African students' actual needs and local opportunities.