WHO Issues AI ethics and governance guidelines for LLM’s

WHO issues ethical guidelines for AI in healthcare: Navigating LMMs for public health enhancement

The World Health Organization (WHO) is poised to release comprehensive guidance on ethical use of AI in healthcare, addressing considerations surrounding Large Multi-Modal Models for public health. The forthcoming guidance encompasses recommendations directed at governments, technology companies, and healthcare providers to ensure the responsible and beneficial use of LMMs for the betterment of public health.

LMMs, distinguished by their capability to process various data inputs such as text, videos, and images, exhibit the ability to generate diverse outputs, mimicking human communication and performing tasks beyond explicit programming. The potential benefits of LMMs in healthcare are significant, spanning applications in diagnosis, patient care, education, and research. However, there are documented risks associated with their usage, including the potential for generating false, inaccurate, or biased information that could adversely impact health-related decision-making.

The WHO guidance addresses five broad applications of LMMs in health and emphasizes the critical importance of transparent information and policies to manage their design, development, and use. One of the central tenets of the guidance is the need to acknowledge and mitigate potential risks, such as biased data and cybersecurity vulnerabilities. Moreover, the guidance encourages active engagement from various stakeholders, including governments, technology companies, healthcare providers, patients, and civil society, in order to foster a collaborative approach to the ethical use of LMMs.

Key recommendations outlined by the WHO include urging governments to invest in public infrastructure to support the ethical development of AI, enforcing laws and regulations to uphold ethical obligations and human rights standards, establishing regulatory agencies specifically tasked with assessing LMMs, and implementing mandatory post-release auditing and impact assessments conducted by independent third parties.

Developers are advised to actively engage a diverse range of stakeholders, including potential users, healthcare professionals, and patients, in the design process of LMMs. This collaborative approach is essential to ensure that the technology is aligned with the needs and values of those it is intended to serve. Furthermore, the guidance emphasizes the importance of tailoring LMMs for well-defined tasks with a focus on accuracy and reliability. Developers are encouraged to anticipate and understand potential secondary outcomes, including unintended consequences, to enhance the overall ethical framework of LMM development.

A significant aspect of the WHO guidance underscores the necessity for cooperative leadership among governments globally to effectively regulate the development and use of AI technologies, particularly LMMs, in the healthcare sector. The global nature of these technologies necessitates a harmonized and collaborative effort to establish standards and guidelines that can be universally adopted.

In conclusion, the WHO's forthcoming guidance on LMMs in healthcare represents a crucial step towards ensuring the ethical and responsible development and use of AI technologies. By addressing potential risks, advocating for transparency, and emphasizing collaboration among diverse stakeholders, the guidance provides a comprehensive framework for navigating the complexities of LMMs in the healthcare landscape. As the field of AI continues to evolve, these guidelines will serve as a cornerstone for shaping a future where technology enhances healthcare outcomes while upholding ethical standards and safeguarding human well-being.