The increasing integration of artificial intelligence (AI) in academic writing has led universities to require AI disclosure in student submissions. This study investigates whether such declarations influence grading and how lecturers perceive AI-assisted work. An experimental study compared grading outcomes between two instructor groups: one assessing an AI-labeled text (n = 15) and another evaluating the same text without AI disclosure (n = 10). Results showed no statistically significant differences in grades. This observation means that, in the conducted experiment, the AI declaration had no impact on lecturers’ evaluations. To further explore lecturers’ perspectives on the use of AI in academic writing, cluster analysis was used to identify three attitudinal groups: 1) AI Skeptics with a focus on responsibility, who support responsible AI use but express moderate concerns; 2) Pragmatic Supporters, who acknowledge both benefits and risks; and 3) AI Critics, who view AI negatively in academic writing.