WHO releases guidelines for multi-modal generative AI in healthcare, resonates with recommendations for other sectors

WHO releases guidelines for multi-modal generative AI in healthcare, resonates with recommendations for other sectors

Subject: S&T

Section: Awareness in IT

Context:

• The World Health Organization (WHO) has released comprehensive guidance on the ethical use and governance of large multi-modal models (LMM) in healthcare.

Large multi-modal models (LLM):

• LMMs, known for their ability to mimic human communication and perform tasks without explicit programming, have been adopted more rapidly than any other consumer technology in history.
• Example: ChatGPT, Bard, Bert and Gemini.
• The risk from LLMs: The generation of false, inaccurate or biased statements, which could misguide health decisions, the data used to train these models can suffer from quality or bias issues, potentially perpetuating disparities based on race, ethnicity, sex, gender identity or age, the accessibility and affordability of LMMs, and the risk of ‘automation bias’ in healthcare, leading professionals and patients to overlook errors, and cybersecurity.

About the Guidelines:

• The guiding document identified five broad applications of LMMs in healthcare: Diagnosis and clinical care, such as responding to patients’ written queries; patient-guided use for investigating symptoms and treatments; clerical and administrative tasks in electronic health records; medical and nursing education with simulated patient encounters; and scientific research and drug development.
• For developers, the WHO advises engaging a wide range of stakeholders, including potential users and healthcare professionals, from the early stages of AI development. It also recommends designing LMMs for well-defined tasks with the necessary accuracy and understanding of potential secondary outcomes.
• It offers a roadmap for harnessing the power of LMMs in healthcare while navigating their complexities and ethical considerations.
• This initiative marks a significant step towards ensuring that AI technologies serve the public interest, particularly in the health sector.
• The six core principles identified by WHO are: (1) protect autonomy; (2) promote human well-being, human safety, and the public interest; (3) ensure transparency, explainability, and intelligibility; (4) foster responsibility and accountability; (5) ensure inclusiveness and equity; (6) promote AI that is responsive and sustainable.

WHO listed out concerns that called for rigorous oversight needed for the technologies to be used in safe, effective and ethical ways. These included:  

• The data used to train AI may be biased, generating misleading or inaccurate information that could pose risks to health, equity and inclusiveness;
• Large language models (LLM) generate responses that can appear authoritative and plausible to an end user; however, these responses may be completely incorrect or contain serious errors, especially for health-related responses;
• LLMs may be trained on data for which consent may not have been previously provided for such use, and LLMs may not protect sensitive data (including health data) that a user provides to an application to generate a response;
• LLMs can be misused to generate and disseminate highly convincing disinformation in the form of text, audio or video content that is difficult for the public to differentiate from reliable health content; and
• Policy-makers should ensure patient safety and protection while technology firms work to commercialise LLMs.

Key recommendations for governments include:

• Investing in public infrastructure, like computing power and public datasets, that adhere to ethical principles
• Using laws and regulations to ensure LMMs meet ethical obligations and human rights standards
• Assigning regulatory agencies to assess and approve LMMs for healthcare use
• Introducing mandatory post-release audits and impact assessments

Increased inequality due to generative AI:

• The 2024 World Economic Situation and Prospects report indicates that while AI’s transformation of the labour market and productivity is underway, there is a concern about its unequal impact.
• The report highlights a global shift toward mass adoption of generative AI technologies, such as ChatGPT, with a third of surveyed firms adopting it within six months.
• However, worries are expressed about AI contributing to increased inequalities within and between countries, potentially diminishing the demand for low-skilled workers and adversely affecting disadvantaged groups, particularly women in roles with higher automation risks.
• Additionally, challenges are emphasized for workers in low-income developing countries, who may experience fewer job disruptions from automation but are also less likely to benefit from AI-driven productivity gains, exacerbated by infrastructure gaps in digital education and internet access.

Disinformation and Misinformation:

• The World Economic Forum’s Global Risks Report for 2024 highlights AI-generated disinformation and misinformation, particularly through deep fakes, as a major global risk.
• The report expresses concerns about potential threats to the legitimacy of newly elected governments, especially during significant upcoming elections worldwide.
• In response to these concerns, the European Union passed the AI Act in December, aiming to ensure the safe and ethical use of AI systems within the EU while respecting fundamental rights.
• The report also identifies quantum computing as a potential disruptor, emphasizing security concerns related to “harvest attacks” on encrypted data for future decryption using advanced quantum computers.

Source: DTE