To a large extent, the focus of contemporary emergency management is on natural, technological, and man-made hazards, such as floods, earthquakes, tornadoes, industrial accidents, extreme weather events, and cyber attacks.
However, with the availability and increasing functionality of artificial intelligence (AI), we may soon see new public safety risks associated with these technologies. It is important to reduce the probability of these risks occurring and be prepared to address them.
Over the past 20 years, my colleagues and I, along with many other researchers, have been using AI to develop models and applications that can identify, assess, predict, monitor, and detect hazards, providing information for emergency actions and decision-making.
We are now at a turning point where AI is becoming a potential source of risk that should be included in risk and emergency management phases - mitigation or prevention, preparedness, response, and recovery.
AI hazards can be classified into two categories: intentional and unintentional. Unintentional hazards refer to those caused by human errors or technical failures.
As the use of AI increases, human errors or AI-based technical failures in AI models will lead to more adverse events. These events may occur in various industries, including transportation (such as drones, trains, or autonomous vehicles), power, oil and gas, finance and banking, agriculture, healthcare, and mining.
Intentional AI hazards refer to the harm caused to personal and property safety through the use of AI. AI can also be used to gain illegal benefits by compromising security systems.
In my opinion, this simple classification of intentional and unintentional hazards may not be sufficient. Here, we need to add a new category of threats - the possibility of AI replacing human control and decision-making. This can be triggered intentionally or unintentionally.
Many AI experts have already issued warnings about such potential threats. Recently, researchers, scientists, and other stakeholders involved in AI development have published an open letter calling for a pause in further AI development.
Public safety and emergency management experts use risk matrices to assess and compare risks. Using this approach, hazards are qualitatively or quantitatively assessed based on their frequency and consequences, and their impacts are categorized into low, medium, and high levels.
Hazards with low frequency and minor consequences or impacts are considered low-risk and do not require additional measures to manage them. Hazards with moderate impacts and frequencies are considered medium-risk and require close monitoring.
Hazards with high frequency or significant impacts, or both, are classified as high-risk. Additional measures need to be taken to reduce and mitigate these risks. Failure to take appropriate action immediately may result in serious personal and property losses.
So far, AI hazards and risks have not been added to risk assessment matrices. We should quickly start incorporating potential AI risks into local, national, and global risk and emergency management.
AI technology is increasingly being applied by institutions, organizations, and companies in various fields, and AI-related hazards are beginning to emerge.
In 2018, KPMG developed an "AI Risk and Control Matrix" that highlights the risks associated with the use of AI by businesses and urges them to recognize these emerging risks. The report warns that AI technology is advancing rapidly and risk control measures must be taken before it goes beyond system control.
Governments around the world have also started developing risk assessment guidelines for the use of AI-based technologies and solutions. However, these guidelines are limited to risks such as algorithmic bias and infringement of individual rights.
At the government level, the Canadian government has issued "Directive on Automated Decision-Making" and established appropriate governance mechanisms to ensure that federal institutions minimize risks associated with AI systems.
The main objective of this directive is to reduce risks to users, federal institutions, and Canadian society when deploying AI systems. According to the directive, each department must conduct a risk assessment to ensure appropriate safeguards are taken in accordance with government security policies.
In 2021, the US Congress tasked the National Institute of Standards and Technology with developing an AI risk management framework for the Department of Defense. The proposed voluntary AI risk assessment framework recommends prohibiting the use of AI systems with unacceptable risks.
At the national level, policy concerns regarding AI are mostly driven by national security and global competition, specifically the risks to national security and economic risks posed by lagging behind in AI technology.
The US National Security Commission on Artificial Intelligence emphasizes the national security risks associated with AI. These risks do not come from the technology itself but from the risk of losing in the global competition in AI development, including with countries like China.
The World Economic Forum highlighted in its "Global Risks Report 2017" that AI is just one of the emerging technologies that could exacerbate global risks. The report considers superintelligent AI systems as a theoretical threat when assessing the risks posed by AI.
However, the latest "Global Risks Report 2023" does not even mention AI and AI-related risks, indicating that the leaders of global companies advising the report did not consider AI as a direct risk.
In terms of understanding, anticipating, and managing risks, the development of AI is far outpacing government and corporate policies. The current global situation, coupled with market competition in AI technology, makes it difficult for governments to have the opportunity to pause AI and develop risk governance mechanisms.
While we should collectively strive for such governance mechanisms, we all need to be prepared for the significant and catastrophic impacts of AI on our systems and society.
Alphabet Q1 paid subscribers reached 350M, fueled by YouTube & Google One. YouTube ad revenue slightly missed at $9.88B amid Premium growth. Total Q1 revenue hit $109.9B, beating forecasts.
Anthropic releases Claude connectors for Adobe, Blender, and Ableton, enabling AI workflow automation and app control. It also sponsors the Blender Development Foundation.
Apple introduces an App Store model splitting 12-month plans into monthly installments. On iOS 26.4+, it lowers upfront costs without reducing total price, adding billing tracking.
Google launches Aadhaar digital IDs in Wallet for India, enabling secure QR/NFC verification. A phased partner rollout targets key sectors, while passport-based credentials expand globally.
YouTube tests "Ask YouTube," an AI tool blending text and video for complex queries. Initially for US Premium users, it supports interactive follow-ups, expanding Google's AI search.
Google grants US DoD classified AI access, joining OpenAI and xAI in defense deals amid Anthropic's dispute. Contractual safeguards against surveillance and lethal weapons remain unenforceable.
Mistral AI launches Workflows, a Temporal-powered orchestration platform for secure, Python-driven enterprise AI automation featuring human-in-the-loop approvals.