Artificial intelligence (AI), and particularly ChatGPT, is increasingly utilized in academia, providing benefits from writing assistance to explaining complex concepts. However, concerns persist about its potential to generate inaccurate or nonsensical information. This paper investigates the strategies students use to collect and validate information from ChatGPT, particularly by differentiating between static and dynamic information. We adopt a qualitative interpretive approach to closely examine the strategies employed by students. The paper concludes with a discussion based on the observed approaches taken by students to collect and validate AI-generated content, highlighting the flaws of these approaches and providing suggestions for AI-enhanced education.