According to TechPinas, students risk damaging their academic credibility when they rely on generative AI to produce citations, as these systems are not connected to real journal databases and frequently invent references that cannot be verified.
The article explains that large language models predict word sequences based on training data, not live academic repositories, making them unreliable for sourcing. Instead, a smarter research workflow combines AI for brainstorming and outlining with trusted academic databases, citation managers, and careful source verification to maintain integrity.
To use AI responsibly, TechPinas recommends crafting prompts that focus on keywords, research strategies, and conceptual explanations, while leaving actual source collection to legitimate scholarly databases. This approach preserves the benefits of AI without introducing fabricated citations into bibliographies.