Faculty Perceptions on Generative AI's Threat to Creativity versus its Potential as a Learning Tool, and the Resulting Need for AI-Resistant Assignment Training: The Pedagogical Divide

General Artificial Intelligence (GenAI), particularly Large Language Models (LLMs), is increasingly being integrated into the higher education system, creating friction within it. However, many reputable institutions in Pakistan and their faculty members have a biased view on the inculcation of GenAI in academics. The main problem arises in disciplines like Business Administration, Information Systems, and Computer Science. This study aims to review this bias using responses from 20 faculty members through in-depth semi-structured interviews. With the use of thematic analysis, the following pedagogical divides were identified: First, the Polarization of Pedagogical Trust, showing problems in academic integrity; Second, the Scaffolding Imperative and Contextual Evasion, identifying hardships in creating AI-resistant assignments; and Third from Instrumentality to Strategic Prompt Competency, highlighting the need for polished frameworks to include ethical integration and prompt engineering. In conclusion, the findings highlight the need for systematic support to help faculty members develop and design step-by-step processes to ensure academic credibility. This way, graduates will be better prepared for the technology-intensive fields that require the use of GenAI.