In July 2024, the National Council on Interpreting in Healthcare (NCIHC) unveiled its working paper, “Guidance For Healthcare Organizations Evaluating the Potential Use of AI-generated Interpreting,” providing essential guidelines for healthcare organizations considering the integration of AI-generated interpreting services.
This new guidance comes on the heels of the release of updated privacy and civil liberties protections and new AI and Machine Translation requirements in the Section 1557 Final Rule, outlining specific operational guidelines for contracting and implementing AI-generated interpreting services. It emphasizes the importance of maintaining high standards in quality, accuracy, and patient safety, highlighting that these services should “support patient safety, good patient outcomes, compliance with legal and accreditation standards, and financial efficiency.”
Building on SAFE-AI's Framework
The NCIHC’s guidance is based upon the foundational framework established by the SAFE-AI Task Force’s “SAFE-AI Guidance on AI and Interpreting Services,” released just a month earlier in June 2024. The SAFE-AI document set forth four key ethical principles: end-user autonomy, safety and wellbeing, transparency, and accountability. It aims to ensure that AI technology enhances rather than compromises the quality of interpreting services, advocating for a balanced and responsible approach to AI deployment. As noted in the NCIHC guidance, these principles are crucial for “providing a standard of service comparable to human interpreters” and ensuring that AI tools are subject to rigorous quality control measures.
NCIHC's Practical Focus
The NCIHC guidance specifically addresses the operational aspects of implementing AI-generated interpreting services in healthcare settings. It offers a series of practical recommendations to ensure these services meet the industry’s high ethical and professional standards. Key recommendations include:
- Informed Consent and User Education: The guidance stresses that patients must be fully informed about the capabilities and limitations of AI-generated interpreting services. It is essential for healthcare providers to offer clear information, allowing patients to make informed decisions and opt-in or opt-out as they see fit.
- Quality Assurance: The NCIHC document underscores the necessity of rigorous quality assurance processes to ensure AI-generated services meet the required standards of accuracy and reliability. This is particularly critical in healthcare settings where precision is paramount.
- Data Privacy and Security: Given the sensitive nature of healthcare information, the guidance emphasizes stringent data privacy and security measures. Healthcare organizations must adhere to relevant regulations, such as HIPAA, to protect patient data.
- Cultural Competency and Accessibility: The guidance also calls for AI tools to be culturally competent and accessible, ensuring they cater to the diverse linguistic and cultural needs of patients.
The Path Forward
The NCIHC’s practical guidance, by building on the SAFE-AI framework, provides healthcare organizations with a simple yet focused roadmap for ethically and effectively integrating AI-generated interpreting services. It highlights the need for ongoing vigilance and adaptation to ensure these services continue to meet evolving standards and patient needs. The document’s focus on practical implementation makes it an invaluable resource for healthcare providers, helping them navigate the complexities of AI integration while safeguarding patient rights and safety.
For more detailed information, access the full guidance on the NCIHC website.
2024 is the Year of AI Transformation
AI is transforming US government agencies, bringing both opportunities and challenges. Addressing ethical, legal, and regulatory concerns is critical to ensure AI’s responsible use.
Talk with an expert at MasterWord about bringing your Language Access Plan to full compliance, including “human review” and “human-in-the-loop” mechanisms to maximize your plan for AI integration: [email protected].