Bold Digital Health

Written Submission Global Dialogue on Artificial Intelligence Governance Submitted by: Rafael M. Lopes, MBA, CPHIMS, CPXP Organization: Bold Digital Health, Los Angeles, California Role: Digital Health Innovation Strategist | Founder Practice: Serving clients across the United States and Latin America Introduction The following recommendations address the organization and content of future sessions, offered through the lens of health equity and grounded in over 25 years of experience in digital health innovation, clinical systems transformation, and health equity strategy. Recommendation 1: Invest in Capacity Building It is difficult for individuals and institutions to shape governance frameworks they haven't had the opportunity to understand. Across much of the world, AI is arriving faster than the capacity to govern it — and this gap spans governments, civil society, and clinical institutions globally. Structured, accessible, multilingual AI orientation materials should be made available before and throughout future sessions — designed for non-technical audiences and framed around practical governance questions rather than technical specifications. Every region, regardless of resources or readiness, must have a meaningful and informed seat at this table. Inclusive participation is not achieved by presence alone. It requires preparation. Recommendation 2: Embed AI Tools Across the Entire Governance Process AI should be embedded across the full lifecycle of this dialogue — to synthesize input, map recommendations against existing frameworks from governments, standards bodies, and professional organizations worldwide, and identify where alignment exists and where conflicts or gaps remain unaddressed. This is work that takes months manually. AI can help humans do it more efficiently and with greater accuracy. AI handles the volume. Humans retain the judgment. The process should be transparent, auditable, and built on openly governed tools. In doing so, this forum becomes a proof of concept for responsible AI use. Recommendation 3: Prioritize Healthcare as a Key Thematic Domain Healthcare is already adopting AI 2.2 times faster than the broader economy, with sector spending nearly tripling in a single year (Menlo Ventures, 2025 State of AI in Healthcare — menlovc.com/perspective/2025-the-state-of-ai-in-healthcare). That pace signals both the transformative potential and the governance urgency. Without intentional governance, AI risks deepening existing inequities in who receives accurate diagnoses and quality care — along racial, economic, and geographic lines. Getting it right improves outcomes for millions globally. Getting it wrong causes measurable harm. Healthcare is not a special interest. It is a leading indicator of what AI governance done well — or done poorly — looks like in practice. It deserves a dedicated thematic focus in Geneva this July. Closing Governance is most effective when it is inclusive, informed, and connected to the real-world consequences of getting it wrong. I am grateful for the opportunity to contribute and look forward to the continued dialogue.

Transparency, accountability, and human oversight;AI capacity-building;Safe, secure and trustworthy AI;Social, economic, ethical, cultural, linguistic and technical implications of AI;

all except open-source items were a priority for me. i only selected the top 4. my next would be interoperability of governance approaches to help cross-border interoperability for a better global cohesion and impact to humans globally. it is often an obstacle if not addressed.

the unique implications of governance in healthcare. deserves some focused attention.

they amplify inequity and prevent realization of health equity. lack of transparency with data models, lack of notification and consent with patients when using ai in healthcare settings, poorly trained AI model that are inherently biased due to stale legacy healthcare data/records that have not been updated to address diverse populations, and lack of capacity building and meeting individuals where they are. these all must be considered in the governance recommendations along with the typical elements of security, safety

interoperability of governance models and frameworks and alignment with as many trusted, expert-led frameworks will allow for improved development of ai solutions that can scale globally and provide max impact across all regions. a global perspective is absolutely necessary

some effort to gather and synthesize governance frameworks from trusted expert sources around the world should be convened to facilitate how to create that global integration with the goal of trying to deliver as much of a common global framework as possible.

Create a multilingual website to capture comments broadly on a website that is promoted widely / globally for that purpose with an memorable and easy to capture and remember domain. then conduct some structured, human-tested and monitored process to aggregate and assemble results in real time so the UN, Unesco, ITU organizers may adapt the programs and dialogs with data

individuals who are not higly literate or have limited access to broadband / internet may be left behind. i recommend some offline/non-digital outreach to some sample communities like this to ask them how to reach similar groups and then conduct research that does not require digital access or visibility

#NAME?

#NAME?