Will Computer-Science Faculty Supply Content to AI Models?

p strongComputer-science faculty often believe their work could help train AI models, but voluntary submission is rare and uncertainty about unauthorized use is widespread./strong/p p Primary Research Group found that 53.63% of respondents considered some of their research or teaching content suitable for AI training. Only 7.25%, however, had actually submitted material for inclusion in a model. Nearly half&mdash43.48%&mdashwere unsure whether their content had already been used without permission, while 14.49% believed unauthorized use had occurred./p p Course-level experimentation is further ahead than formal institutional programs. About 23.19% had placed their own notes, lectures, videos or research into a chatbot or model for students. By contrast, just 4.35% said their department had made an effort to collect and prepare content for AI use. Faculty initiative is therefore developing faster than shared policy or infrastructure./p p Universities should separate three decisions that are easily blurred: using faculty material in a closed course tool, licensing content to an external developer, and releasing content for broader model training. Each raises different questions about ownership, student information, third-party rights, attribution, compensation, retention and withdrawal./p p The immediate management need is a clear pathway. Faculty should know who can review a proposal, what permissions are required and what contract terms protect their work. Without that pathway, potentially useful collaborations may stall while informal experiments proceed with less oversight./p p  /p