A new study, based on analysis of online discussions in the CrimeBB database, finds that many cybercriminals appear to be struggling to adopt artificial intelligence in their work. Researchers reviewed posts and conversations where participants discuss using AI tools, capabilities, and workflows for cyber activities. Across the discussions, the research suggests that while interest in AI is present, practical adoption is limited by obstacles such as knowledge gaps, technical difficulties, and uncertainty about how to integrate AI into existing processes. The study characterizes the situation as uneven, with some participants exploring AI-related methods but many expressing challenges or confusion about implementation. Rather than indicating widespread, effective use of AI by criminals, the findings point to a learning curve and constraints in turning AI into usable tools for cybercrime. The results reflect how rapidly emerging technologies do not automatically translate into operational advantage, especially for users who may lack the necessary expertise or resources.