Responsible AI Use in Accredited Continuing Education: A Comprehensive Guide

Artificial Intelligence presents a powerful opportunity to enrich accredited CE programs. When employed thoughtfully, AI tools can serve as a valuable adjunct to educational design, augmenting human insight and professional judgment rather than replacing it. Accredited providers are encouraged to experiment with AI to enhance various aspects of their educational offerings. This includes utilizing AI for comprehensive professional gap and learning needs assessments to identify emerging trends or at-risk populations. In content design and development, AI can assist with brainstorming, drafting, and adapting materials for diverse audiences. Furthermore, AI can aid in analyzing financial relationships with ineligible companies to ensure content relevance and independence. For assessment, AI can generate new assessment methods, create distractors and answer rationales, and develop scoring rubrics. AI also offers capabilities for summarizing feedback and analyzing large evaluation datasets, supporting quality improvement initiatives, and facilitating social media outreach. Finally, AI can be instrumental in building complex single-profession and interprofessional case studies and scenarios in intricate clinical domains.

A fundamental requirement for accredited CE is the safeguarding of independence and the mitigation of bias, a principle that extends rigorously to the use of AI. Accredited Providers must ensure that AI-generated or AI-assisted content does not introduce bias or undue influence. All content must strictly adhere to the Standards for Integrity and Independence in Accredited Continuing Education, meaning it must be based on current science, evidence, and clinical reasoning, presenting a balanced view of treatment and diagnostic options. Crucially, no aspect of AI-driven content should reflect the interests of ineligible companies or incorporate promotional language. This includes ensuring that advertisers, sponsors, or commercial supporters have no influence over AI-driven recommendation engines used in accredited activities. The provider must retain full control and responsibility for all educational content, with decisions regarding AI use made by individuals free from relevant financial relationships. The same stringent standards of content review and independence applied to human-authored materials must be maintained for AI-generated content. This involves actively screening AI outputs for commercial bias or promotional language that could benefit ineligible companies and meticulously documenting the steps taken to ensure AI-generated content is free from such influences.

While AI offers significant enhancements in efficiency and design capabilities, it is imperative to recognize that it is a tool and not a substitute for professional judgment. Current AI systems are not consistently capable of making ethically sound or context-specific decisions without human oversight. Generative AI, in particular, is prone to fabricating citations or producing misleading content, and inherent inaccuracies or training biases can lead to harmful or discriminatory outputs. Therefore, content integrity and data stewardship remain fundamentally human responsibilities of the accredited provider. For AI-generated materials that are fixed—meaning they are created once and then distributed, such as slides, handouts, or assessments—the accredited provider must ensure these outputs are thoroughly reviewed and approved by named, qualified individuals before dissemination. This review process must include checking for factual errors or 'AI hallucinations,' and screening for bias or stereotyping in clinical or demographic representations. Version control and traceability, indicating who reviewed what and when, are also essential. Experienced clinicians must provide oversight, confirming that all recommendations and outputs are accurate, scientifically valid, unbiased, and reliable. When learners use AI systems in real-time to answer clinical questions or generate dynamic outputs, the AI system and its underlying processes must undergo rigorous oversight by identified, experienced clinicians to ensure accuracy, validity, and lack of bias. Learners should be reminded that AI systems can err and that ultimate responsibility for clinical decisions rests with the human professional.

Certain applications of AI carry inherent risks that necessitate careful consideration and, in many cases, prohibition or rigorous oversight within accredited Continuing Education. AI should generally not be used for generating diagnostic or treatment recommendations without thorough clinical validation. Similarly, entering or storing sensitive content in public or non-secure AI platforms is strongly discouraged due to potential breaches of confidentiality and security. The auto-production of assessment answers that are visible to learners is also prohibited, as it undermines the integrity of the assessment process. Automating analysis or summaries that bypass essential bias, independence, or accuracy checks is another high-risk use case that must be avoided. If any of these high-risk applications are considered, they must be accompanied by exceptionally rigorous oversight and expert review to mitigate potential harms and maintain the standards of accredited CE.

Exercise caution when uploading educational materials into AI systems, especially if the content includes patented, proprietary, or unpublished materials from faculty. Obtaining explicit permission from faculty is paramount, particularly when using open-access or third-party AI platforms, to avoid exposing trade secrets or violating confidentiality agreements. To protect data integrity, provider credibility, and adherence to accreditation standards, accredited providers should exclusively input confidential or proprietary information into 'private' or 'closed' (company-specific/licensed) AI systems. Providers must monitor and comply with all state and future legal or regulatory requirements related to AI, implementing best practices and anticipating future laws. AI-assisted content and learner data should only be stored in secure, access-controlled databases compliant with institutional and legal data protection standards (e.g., HIPAA, FERPA, GDPR). AI systems and Large Language Models (LLMs) should be hosted in secure, vetted environments approved by the institution’s IT or compliance office. Routine audits are necessary to ensure AI-generated or processed content does not violate content validity, independence, or learner confidentiality requirements. The use of free or open-access AI platforms for accreditation-sensitive tasks should be avoided unless explicitly permitted by the organization and accompanied by sufficient risk mitigation. Ultimately, the use of AI systems must align with existing standards of integrity and freedom from commercial influence required for accredited CE activities.