Generative artificial intelligence technologies have revolutionized the research landscape, with significant implications for Digital Humanities, a field inherently intertwined with technological progress. This article investigates how DH scholars adopt and critically evaluate generative AI technologies such as ChatGPT in research. Drawing on 76 responses collected from an international survey study, we explored DH scholars' rationale for adopting or not adopting generative AI tools in research, identified the specific practices of using generative AI tools to support various DH research tasks, and analyzed scholars' collective perceptions regarding the benefits, risks, and challenges of using generative AI tools in DH research. The survey results reveal two key findings: first, DH research communities hold divisive opinions about the value of generative AI in DH scholarship; second, scholars have developed new practices and perceptions for using generative AI tools, which differ from those associated with traditional AI-based tools. Our survey represents one of the first survey-based analyses on this topic. It has the potential to serve as a building block for future empirical inquiries into the impact of generative AI on DH scholarship.