This study explores how knowledge workers perceive the influence of external generative AI tools on knowledge creation processes. While most existing research has focused on integrated AI tools within organizational systems, this study addresses the gap in understanding how external generative AI tools like ChatGPT are individually used and perceived by employees. Through a qualitative multiple-case study design, data were collected from nine knowledge workers across two Swedish organizations through semi-structured interviews. Thematic analysis was conducted using the SECI model of knowledge creation as a theoretical lens. Findings reveal that external generative AI tools can positively influence knowledge creation by supporting brainstorming, enhancing communication, and increasing efficiency. However, the study also identifies risks related to data security, trust in AI-generated content, and its impact on collaboration and communication. Differences between the two organizations suggest that organizational culture, formal training, and support play a key role in shaping perceptions and usage of these tools. The study contributes to research on generative AI in knowledge management by examining how unregulated use of external generative AI interacts with established knowledge creation models and highlights the need for clearer policies and guidance.