P248: Copyright Compliance in Artificial Intelligence (AI)-Assisted Medical Information (MI) Content Generation
Poster Presenter
Namareq Musa
Post-Doctoral Fellow
Pfizer United States
Objectives
To evaluate the accuracy and practical utility of AI-assisted medical content generation within MI by examining how access to copyright-protected materials affects output quality and compliance, and to determine whether AI can reliably support MI teams in producing evidence-based content.
Method
A survey was distributed to the Pharma Collaboration for Transparent Medical Information (PhactMI) members to assess AI adoption, copyright impact, and practices in MI workflows.
Results
Survey results revealed limited organizational infrastructure and training related to copyright and AI use. Only 27.9% of respondents reported that their company has a dedicated copyright liaison for medical information, while 66.7% indicated no such role. Formal training on copyright compliance when using AI within the past 12 months was reported by 38.9% of respondents, whereas 61.1% had not received training. Use of AI to summarize or generate SRDs varied, with 22.2% reporting current use, 22.2% reporting no use, and 55.6% selecting other. Despite these gaps, respondents demonstrated interest in further engagement, with 27.8% expressing openness to participating in a working group or roundtable discussion. When asked about helpful resources, respondents most frequently identified copyright decision trees and case studies (both 61.1%), followed by sample guidance documents and training modules (both 55.6%).
Conclusion
AI use in MI content creation is rapidly increasing, but readiness varies widely, underscoring the need for practical tools and structured training to ensure compliant and responsible AI-supported content creation.
