Global Shield Briefing (11 September 2024)
Global Shield Briefing (11 September 2024)
The governance of artificial intelligence, food security and global catastrophic risk
The latest policy, research and news on global catastrophic risk (GCR).
You look out from the ship’s bridge, scanning the stormy seas, seeking the horizon hidden by ominous storm-clouds, as waves crash against the hull. You must navigate through these turbulent times. Prepared for rough conditions, you adjust course and marshal your resources, knowing full-well that your ship can weather the unexpected. You direct with purpose, confident in your people, strategy, and processes. To not only survive. But to emerge stronger and steadier for whatever lies beyond. Risk and uncertainty are no match for a well-governed vessel.
Handling the early phase of AI governance
China’s government is taking catastrophic AI risk more seriously, according to a detailed article by Carnegie Endowment researcher, Matt Sheehan, and another article in The Economist. The official readout of the Third Plenum of the Communist Party of China’s Central Committee noted that the committee agreed to “establish an AI safety regulatory system”. Interestingly, this commitment was placed in a section on “Improving the public security governance mechanisms”, alongside major public emergencies, natural hazards, food safety and security, and biological hazards. An op-ed in the People’s Daily in August provided further semi-official explanation on the need for AI safety and governance.
In the US, AI companies are being critiqued for not taking AI safety seriously enough. A RAND commentary notes that safety is not a priority. A piece in The Atlantic states that “novel corporate-governance structures cannot constrain executives who are hell-bent on acceleration.” In The Nation, Lawrence Lessig, co-founder of the nonprofit Change Congress, looks at how big tech companies have lobbied heavily against the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act (SB1047). SB1047 requires that those developing the most advanced AI models actively reduce or avoid the risk of causing “critical harm”. It has passed the Californian legislature but may still be vetoed by Governor Newsom.
MIT FutureTech and its collaborators have released a comprehensive review of AI risk taxonomies, finding over 700 individual ‘risks’, which can be categorized by 7 domains and 23 subdomains.
Policy comment: The governance of AI is at a critical, and tenuous, juncture. Beijing and Washington – as well as Sacramento – are taking initial steps to govern AI and promote AI safety. Whether both countries deliver on these early commitments is highly uncertain, especially given the strategic competition between them. As Sheehan states in his piece, “if extraordinarily powerful – and potentially dangerous – AI systems are on the technological horizon, the competitive pressure to be the first country to build them will be immense.” How the US and China develop AI capabilities for security purposes, governs AI safety in the private sector, and engage with each other on AI risk, will influence how the rest of the world develops, uses and regulates AI. At this early stage of AI policy development, governments should see AI risk in its totality by considering its societal, economic, security, environmental, infrastructure and ethical challenges.
Also see:
A note by the Civilization Research Institute on how AI might cause more problems for climate change than it might solve.
An interview with Yoshua Bengio, a leading voice on AI risk and one of Time100 AI 2024, on the need for AI regulation.
A call for submissions for an Australian government consultation on introducing mandatory guardrails for AI in high-risk settings, together with their first iteration of the Voluntary AI Safety Standard.
Leading on global food supply security
.jpg){kind=link}
A new report by the Atlantic Council looks at Brazil’s important role in global food security out to 2050. It notes that “Brazil, the United States, and other like-minded partners must work together to ensure that global food supply grows to meet rising demand in the future in ways that are both environmentally and economically sustainable.”
Brazil’s President Lula has put food security for hunger and poverty reduction as a key element of Brazil’s global leadership agenda for the G20 in 2024 and COP30 in 2025. In July, Lula announced the Global Alliance Against Hunger and Poverty, with the official launch scheduled during the G20 Leaders Summit in November. More than 100 countries are expected to join.
Policy comment: Brazil is uniquely positioned to lead on global food supply security. It is one of the world’s largest agricultural producers and exporters, including being China’s primary food supplier. And it is one of the world’s leaders on food and agricultural research and development; in the Global Food Security Index 2022, Brazil was rated 100/100 on commitment to innovative technologies. However, there is more Brazil could do to fulfil a critical leadership role on global food supply in a crisis. While Lula’s food security initiatives are laudably focused on hunger and poverty reduction, they also need to consider how Brazil will manage catastrophic threats to the food supply. Every major agricultural producer of the world must develop policy mechanisms to manage catastrophic threats, ranging from developing alternative and resilient foods to collaborating internationally with other nations on plans and shared resources, such as the Svalbard Global Seed Vault.
Also sees:
Recent news on the severe natural disasters hitting Brazil. Almost 60 per cent of the country – equating to about half the size of the US – is under stress from drought, the worst in seven decades. Since the start of 2024, Brazil has registered almost 160,000 fires, mostly due to deforestation and clearing for agricultural purposes. Some parts of the Amazon River are at their lowest level on record, which is also impacting soy and corn production in critical agricultural states like Mato Grosso. Major rain is not expected until October.
Governing risk to avert catastrophe
Global Shield has released a report on how governments can better govern global catastrophic risk. For most countries, risk governance is missing or lacking for national-level risk, and, by extension, global catastrophic risk. Risk governance encompasses the organizational functions of government that guide, coordinate and oversee risk reduction efforts. It ensures that risk management across government is appropriately organized, structured, managed, resourced and delivered.
Policy comment: Governments must establish holistic risk governance for national risk. It does not reduce risk directly – ultimately, governments must take actions that prevent and prepare for risk. But risk governance ensures that assessment, prevention and preparedness activities are conducted in a strategic, coordinated and effective manner. Treating risk in departmental or threat-specific siloes can lead to poor prioritization, missed policy opportunities, disjointed implementation and competition over budget and attention.
Risk governance sits across five areas. First, policy guidance and strategic planning. Governments must set a clear direction for reduction of national risk. Second, architecture and institutional design. Governments must establish effective and empowered structures for managing national risk, including GCR. Third, decision-making processes. Governments need to implement decision-making processes that enable quick and decisive action in the face of national risk and GCR. Fourth, risk financing and budgeting. Governments must incentivize risk reduction and ensure risk reduction is adequately financed. Finally, personnel and culture. Governments should enhance the mindset and capabilities of policymakers towards risk reduction.
For those with busy diaries, check out our article in The Bulletin of Atomic Scientists on this topic.
This briefing is a product of Global Shield, the world’s first and only advocacy organization dedicated to reducing global catastrophic risk of all hazards. With each briefing, we aim to build the most knowledgeable audience in the world when it comes to reducing global catastrophic risk. We want to show that action is not only needed, it’s possible. Help us build this community of motivated individuals, researchers, advocates and policymakers by sharing this briefing with your networks.