AIM: AI Incidents and Hazards Monitor

The event involves an AI system (the Mythos model) whose development and use are under scrutiny due to potential cybersecurity risks. While no direct harm has been reported, the article highlights credible concerns from government and financial authorities about possible future harms, including risks to cybersecurity and supply chain security. This fits the definition of an AI Hazard, as the AI system's use could plausibly lead to significant harms, but such harms have not yet materialized. The article focuses on the potential risks and ongoing discussions rather than actual incidents or realized harm.[AI generated]

The article explicitly involves AI systems (chatbots powered by large language models) whose use has directly led to problematic medical advice that could cause injury or harm to users' health, fulfilling the criteria for an AI Incident. The harm is realized or ongoing, as the chatbots are widely used by adults for health queries, and the study documents a high rate of problematic responses that could mislead users. The event is not merely a potential risk (hazard) or a response/update (complementary information), but a clear case where AI use has caused or is causing harm, justifying classification as an AI Incident.[AI generated]

The AI system 'Mitos' is explicitly described as an autonomous agent capable of identifying and exploiting security vulnerabilities, which directly relates to AI system involvement. The article focuses on the plausible risk that misuse of this AI could cause major disruptions to financial infrastructure, a critical infrastructure harm category. Since the harm is not yet realized but the risk is credible and significant, this event fits the definition of an AI Hazard rather than an AI Incident. The article also details responses by financial authorities, but the main focus is on the emerging threat rather than a response to an incident. Therefore, the classification is AI Hazard.[AI generated]

The AI system generated inappropriate content that bypassed safety controls, directly leading to harm in the form of unsafe and non-compliant content dissemination. The company's response and remediation efforts are complementary information but do not negate the fact that the AI system's malfunction caused harm. Therefore, this qualifies as an AI Incident due to realized harm from the AI system's outputs.[AI generated]

The article explicitly states that the job cuts are linked to stepping up AI use and automating manual tasks, indicating the AI system's use is directly leading to harm in the form of job losses. This fits the definition of an AI Incident because the development and use of AI systems have directly led to harm to a group of people (employees losing jobs).[AI generated]

The AI systems involved are large language models used as medical diagnostic chatbots, which clearly qualify as AI systems. The study shows that their use leads to a high rate of diagnostic errors, which can cause harm to patients' health by misguiding treatment decisions. This constitutes direct harm to health (harm category a) caused by the AI systems' outputs. Therefore, this event qualifies as an AI Incident due to the realized harm from the AI systems' use in medical diagnosis.[AI generated]

The event involves an AI system explicitly mentioned (Google's Gemini and other AI tools) used to generate political content that has a large social impact. The AI-generated avatar influences public opinion and election-related discourse, with risks of misinformation and confusion, which are harms to communities and potentially violations of electoral laws. The article details realized social and political harms, not just potential risks, and discusses challenges in accountability and regulation. This fits the definition of an AI Incident because the AI system's use has indirectly led to significant harm to communities and possible legal violations in the electoral context.[AI generated]

The event involves AI systems (ChatGPT, Gemini) making decisions that directly impact people in areas like finance and trust, with documented biases leading to differential treatment based on demographics. This constitutes a violation of rights and harm to individuals/groups due to biased AI judgments. Since the harm is realized and linked to AI system use, this qualifies as an AI Incident under the framework, specifically under violations of human rights or breach of obligations protecting fundamental rights.[AI generated]

The article explicitly states the use of generative AI to create fake biometric data (deepfakes) to circumvent security systems, which directly enabled criminal activities resulting in financial theft and fraud. This constitutes direct harm caused by the AI system's use. Therefore, this qualifies as an AI Incident under the framework, as the AI system's use directly led to violations of property rights and harm to consumers and the company.[AI generated]

The article describes Uber's planned large-scale deployment and investment in autonomous driving technology, which involves AI systems. While no harm has yet occurred, the widespread use of autonomous vehicles could plausibly lead to AI incidents such as accidents or disruptions. Therefore, this event represents a plausible future risk related to AI systems, qualifying it as an AI Hazard rather than an incident or unrelated news.[AI generated]

The article explicitly mentions AI systems in the context of AI-generated deepfake videos and fake news that could undermine election integrity. However, it does not describe any realized harm or incident caused by AI misuse; rather, it is a warning and policy announcement about preventing such harms. Therefore, this event represents a plausible future risk of harm from AI misuse in elections, qualifying it as an AI Hazard. It is not an AI Incident because no harm has yet occurred, nor is it Complementary Information since it is not an update on a past incident but a new government statement about potential risks and responses.[AI generated]

The event involves the use of an AI system (deepfake technology) to generate fake legal documents used in a fraudulent scheme. This use of AI directly led to an attempted harm to a person (financial fraud and potential legal consequences). The harm was averted, but the AI system's role in the fraudulent attempt is clear and pivotal. Therefore, this qualifies as an AI Incident because the AI system's use directly led to an attempted harm, even if the harm was prevented.[AI generated]

The article explicitly involves AI systems (large language models) used in medical diagnosis. It discusses their use and limitations, focusing on their failure to perform initial diagnostic reasoning without human supervision. Although no actual harm (such as misdiagnosis causing injury) is reported, the study warns that unsupervised use could plausibly lead to harm in clinical settings. This fits the definition of an AI Hazard, as the AI systems' malfunction or misuse could plausibly lead to injury or harm to patients if deployed without human oversight. Since no realized harm is described, it is not an AI Incident. The article is not merely complementary information because it centers on the risk and performance limitations of AI in diagnosis, not on responses or ecosystem updates. Therefore, the correct classification is AI Hazard.[AI generated]

The article describes the potential future use of AI systems in policing to assist with child sexual abuse investigations. While AI involvement is clear, the use is still under consideration and not yet implemented, so no harm has materialized. The AI's role could plausibly lead to benefits or risks in victim identification and safeguarding, fitting the definition of an AI Hazard. There is no indication of realized harm or incident, nor is the article primarily about responses or updates to past incidents, so it is not an AI Incident or Complementary Information. It is not unrelated because AI is central to the discussion.[AI generated]

An AI system was involved as the image was generated by AI. The use of this AI-generated image without proper verification led to the spread of false information on a public broadcast, which is a harm to communities by spreading misinformation. The harm is realized, not just potential, as the false image was aired and discussed as if real. Therefore, this qualifies as an AI Incident under the definition of harm to communities caused directly or indirectly by an AI system's output.[AI generated]

The article clearly involves AI systems through the discussion of AI-driven data centers and their resource consumption. While no direct harm has yet occurred, the article warns of potential environmental harm (water scarcity) caused by the growth of AI infrastructure. This fits the definition of an AI Hazard, as the AI systems' use could plausibly lead to harm to communities and the environment. There is no indication of a realized incident or harm, so it is not an AI Incident. The article is not merely complementary information since it focuses on the risk and impact of AI system use on natural resources, not on responses or updates. Hence, AI Hazard is the appropriate classification.[AI generated]

The article involves AI systems in the context of cybersecurity threats and defense, highlighting the potential for AI-enabled cyberattacks and the challenges they pose. However, it does not describe any actual AI-related harm or incidents occurring at present. Instead, it provides a warning and assessment of plausible future risks and challenges associated with AI in cybersecurity. Therefore, this qualifies as an AI Hazard, as it plausibly could lead to AI Incidents but no incident has yet occurred or been reported.[AI generated]

The article centers on the plausible risk that AI systems could be exploited to leak sensitive information covertly, which could lead to significant harm such as intellectual property theft and security breaches. It discusses research that proposes detection frameworks and calls for legal and governance upgrades to address these risks. Since no actual data leakage or harm has been reported, but the risk is credible and the article emphasizes the need for preventive measures, this qualifies as an AI Hazard rather than an AI Incident or Complementary Information. It is not unrelated because it clearly involves AI systems and their potential misuse.[AI generated]

The AI cameras are explicitly described as AI systems designed to detect specific driver behaviors that are among the leading causes of fatal and serious injury collisions. Their use directly influences enforcement actions (fines, penalty points) and aims to reduce harm to people on the roads. Since the AI system's use is directly linked to preventing injury and death, this qualifies as an AI Incident under the definition of harm to health of persons resulting from the use of an AI system.[AI generated]

The event involves an AI system explicitly used by the tax authority to detect fraud, which is a violation of legal and financial rights and harms public property (state revenue). The AI system's use in identifying irregularities and reporting them to authorities directly contributes to preventing or addressing these harms. Since the article describes the AI system's active role in detecting fraud (harm that has occurred or is ongoing), this qualifies as an AI Incident. The harm is related to violations of obligations under applicable law protecting financial rights and property (public revenue).[AI generated]