ICH

First a commitment from member states to adhere to some AI regulation principles across the board. Think EU AI act and GDPR but a the global level. You don't need to reinvent the wheel just bring everyone on the same boat. Second a clear understanding of the cost of inaction, ie What would happen if there is no outcome of the Global Dialogue what is the current cost (economic, human, etc..) of the lack of regulation and global agreements. Third bring the voice of the unheard to the center of the debate which is monopolize by the tech industry.

• Open-source software, open data and open AI models
• Safe, secure and trustworthy AI
• Transparency, accountability, and human oversight
• Interoperability of governance approaches

Right now use of AI in drug development for human use as well as AI assisted therapies is not being regulated. Medicamnt regulatory agencies are lagging behind on catching up with the evolution of the technology ( as many others) there is a need of establishing criteria's and approaches that can help those authorities to evaluate and assess the risk for people. Now if AI models aren't transparent and auditable if they aren't trust worthy how can they be used to develop and approve something as fundamental as medicaments? Yet, this is being done and very very few if any understands the potential consequences.

The environmental impact of AI is nowhere listed here. That's very disappointing! Also Data privacy and IP protection should be an item on its own to be honest.

I highlighted this in my previous answer but basically lack of understanding, accountability on the use of AI in the development and approval of medicaments is a huge problem with potentially huge consequences. Lack of trustworthy and data safe AI models can have great consequences on IP and medicaments development and approval. Finally the lack of accessibility of AI tools creates huge disparities between those who can use them and take advantage of those tools ( i.e to fast track med development ) va those who can't.

Bring everyone to commit to global regulation.

EU AI Act.

Use crowdsourcing tools ( AI toosl) Not this crappy MS form.. really, for a team looking at AI this is a bit embarrassing… And allow for Hybrid participation.

Who is included? I think that's a thematic question, you need to look a it per sector. I.e in my area : patients, doctors, regulators basically everyone except pharma companies that only care about IP protection…

Look this AI Dialogue format is very old school, bringing everyone to the same location do a big conference.. looks fancy, an people love to travel and get pay DSA for that but is a) not inclusive ( many people can't afford to travel b) extremely inefficient. So if you really want to be innovative move away from this idea and create an asynchronous continuous mechanism for engaging everyone. The Dialogue cannot happen once a year in a random location around the world cause a member state wants to boost their ego and make their hotels and restaurants earn some money.. Even the name doesn't make sense a Dialogue is a continuous process no something that happens once a year.

AI Eu Act.