Foresters

1) A commitment to have an international treaty signed by all countries in ensuring that AI use cases that could lead to mutually assured destruction be initiated, adopted, then ratified by all countries (MAD agreement) 2) A commitment to have systemic use of AI infringing on human rights in a given or multiple countries be enforceable, remediated, and brought to the international court of justice. 2) a commitment to ensure AI capacity, skills and knowledge is shared cross border to international interests for the good of humankind rather than nationalized.

• Safe, secure and trustworthy AI
• Protection and promotion of human rights
• AI capacity-building
• Transparency, accountability, and human oversight

A race to nationalize AI economic returns could lead to a race to the bottom from a governance perspective in private sector, allowing negligent or reckless businesses to cut corners on necessary testing and ongoing oversight measures and controls, as well as allow for unethical or undesirable use cases. This can lead to unsafe and vulnerable AI software and hardware, and can lead business decisions to discard human rights impacts such as illegal discrimination in decisions impacting individuals for the sake of the race for profits or survivability. Capacity to build AI should be shared internationally under a principled-based treaty enforceable globally with clear consequences or sanctions against infringing nations. Ultimately, further clarity and appropriate remediation should remain with humans not the AI software or hardware used.

1) Yes, primary focus for AI is typically on software capabilities not hardware capabilities (eg. Chips). Supply chain issues as well as hardware vulnerabilities should be factored into the governance frameworks. 2) Additionally, resources required, ownership and control, as well as energy sources and lasting impact on sustainability relative to the merge footprint, and social, economic and environmental impacts (ESG), should be further explored, as some nations may misleadingly shift their policies under the guise of sustainability to bolster energy availability to AI business, or increase financial pressures on citizens for same.

I note enforceability issues relative to human rights when AI is involved, poor procedural protections to individuals given the asymmetry of power and asymmetry of information in the hands of businesses relative to the use of AI and potential stealth impact on such individuals which would warrant an amendment to existing civil procedure laws (eg. Presumption), as well as potential labour force impact with asymmetrical access to knowledge and resources relative to AI, which feeds the governance gap issue.

See commitment propositions in first answer under this survey.

Representative samples of various sectors, public sector and private sector, education, as well as the general public would allow for better dialogue. Consider a youth forum as well, to build for the future. Consider MILA, Vector institute, Montreal Ai Ethics Institute, as well as ongoing R&D in academic circles Also consider the military sector, as they design and build under different international laws and codes.

Youth is underrepresented. ML software engineers should also be brought into the conversations as they intrinsically know the technical limitations and potential benefits of AI. Technology philosophy or philosophy or sciences would allow for a timely international conversation on how to assess and value morals and ethical use cases of AI and build for now and the future.

See the Global Financial Innovation framework initiative asan example. International shared access to AI governance and knowledge would reduce the race to the decisive strategic advantage, allowing for a give-give approach without excessive economic assymetry while still allowing for national economic and competitive advantages. This is only possible if geopolitical tensions are diplomatically de/escalated for global consensus that an international approach to AI design and adoption is most beneficial in the short to long-term.

See MIT risk tool for risk identification. See resources with the Montreal AI Ethics Institute. See the "Social Contract for AI" written by Abhishek Gupta and Mirka Snyder Caron.