AIM: AI Incidents and Hazards Monitor

The article highlights a credible future risk where AI could be used maliciously to create biological weapons, which would cause harm to health (a). The event involves AI system development and use potentially enabling such threats, but no incident of harm has yet occurred. The call for safeguards is a governance response to this plausible risk. Therefore, this is best classified as an AI Hazard, as it concerns plausible future harm from AI-enabled misuse in biotechnology.[AI generated]

An AI system was explicitly used to generate a fake voice recording falsely attributed to a public figure, which was then disseminated widely, causing misinformation and social harm. This constitutes a violation of rights and harm to communities through the spread of false information. The event describes realized harm from the AI-generated content, not just a potential risk, and the police arrest indicates the seriousness of the incident. Therefore, this qualifies as an AI Incident.[AI generated]

The use of generative AI to create falsified hospital documents that enabled repeated fraudulent insurance claims constitutes direct harm caused by an AI system's misuse, fulfilling the criteria for an AI Incident due to financial harm and legal violations. The subsequent government response to establish a task force to prevent AI-enabled insurance fraud is complementary information enhancing understanding of societal and governance responses. Since the main event involves realized harm caused by AI misuse, the classification is AI Incident.[AI generated]

The event involves the use of AI systems leading to realized harm—job cuts in the tech sector—due to AI-driven automation or restructuring. Job loss is a significant harm affecting individuals and communities, fitting within the scope of harm to people (a). Since the harm is occurring and AI is explicitly cited as the cause, this qualifies as an AI Incident rather than a hazard or complementary information.[AI generated]

The article discusses the plausible future environmental harms caused by the development and use of AI systems, specifically through the increased resource consumption of data centers supporting AI workloads. Although no direct harm has yet occurred, the projected doubling of power and water consumption and associated environmental impacts constitute a credible risk of harm. Therefore, this event qualifies as an AI Hazard because it describes a plausible future harm linked to AI system deployment and infrastructure expansion, but no actual incident of harm has yet materialized.[AI generated]

The Grok chatbot, an AI system, was used to create non-consensual deepfake images of MP Jess Asato, which is a direct misuse of AI-generated content causing harm to her privacy and personal rights. The lawsuit alleges violations under data protection laws and privacy invasion, which are clear harms under the AI Incident definition. The involvement of the AI system in generating harmful content and the resulting legal action confirm this as an AI Incident rather than a hazard or complementary information.[AI generated]

The event involves the use of AI systems (large language models) in the development and operation of autonomous computer worms that can self-replicate and adapt to exploit network vulnerabilities. The researchers' demonstration confirms the AI system's role in enabling a new type of cyberattack that could disrupt critical infrastructure and cause harm to property and communities. Although the study was conducted in a controlled lab setting, the plausible and credible risk of real-world deployment and resulting harm qualifies this as an AI Hazard. However, since the AI-driven worms have not yet caused actual harm outside the lab, it is not an AI Incident. The article primarily warns about the potential for harm and calls for preparedness, fitting the definition of an AI Hazard.[AI generated]

The article clearly involves AI systems, such as ChatGPT and AI-powered cloud infrastructure, and discusses their resource consumption and environmental impact. However, it does not describe any actual harm or incident that has occurred due to AI use or malfunction. Instead, it focuses on the potential for significant ecological harm in the future if current trends continue unchecked. This fits the definition of an AI Hazard, as the development and widespread use of AI systems could plausibly lead to environmental harm. The article also includes expert opinions and recommendations for governance, but these are part of the hazard context rather than complementary information about a specific incident or hazard event.[AI generated]

The use of AI to create fake videos and voice impersonations directly leads to harm by enabling a scam that deceives people and attempts to steal their personal information. This constitutes a violation of rights and harm to individuals, fitting the definition of an AI Incident because the AI system's use is central to the harm occurring. The harm is realized, not just potential, as victims are targeted and deceived.[AI generated]

The Vepr robot is an AI-enabled system performing autonomous or semi-autonomous tasks in a combat environment, including medical evacuation and combat support. Its deployment and use have a direct impact on human health and safety by evacuating wounded soldiers and supporting military operations. Since the robot's use directly leads to harm prevention and life-saving, and it operates in a high-risk environment, this qualifies as an AI Incident due to the direct involvement of an AI system in affecting human health and safety in combat.[AI generated]

The Sanctum C-UAS system includes AI components such as battle management software that processes sensor data to detect, track, and engage drones autonomously or semi-autonomously. The event reports a successful interception of a hostile drone, indicating realized use of AI in a defense operation. Although no harm is reported, the system's use in military applications involving lethal force and autonomous targeting qualifies as an AI Incident due to the direct involvement of AI in potentially harmful military actions. The event is not merely a product announcement but a demonstration of AI-enabled lethal capability with direct implications for harm potential in conflict scenarios.[AI generated]

The article explicitly mentions the use of an AI system (live facial recognition) by police to identify individuals at a public event. The system's use is intended to detect alleged criminals, which involves AI system use. However, the article does not describe any actual harm or incident resulting from this deployment, such as wrongful identification, privacy breaches, or other violations. Civil rights groups' concerns about privacy breaches indicate plausible future harm. Since no harm has yet materialized but the AI system's use could plausibly lead to violations of rights or other harms, this event fits the definition of an AI Hazard rather than an AI Incident or Complementary Information.[AI generated]

The article explicitly mentions AI-generated emails, fake digital identities, and synthetic media used by fraudsters to deceive businesses, causing direct financial harm. The AI systems' use in creating convincing fake documents and communications is central to the fraud incidents described. The harm is realized (financial losses, delayed payments, and damage to business operations), meeting the criteria for an AI Incident. The article also discusses mitigation and prevention measures but the primary focus is on the actual harm caused by AI-enabled fraud, not just potential or complementary information.[AI generated]

The article explicitly discusses AI systems (Autobrains' autonomous driving software, Nvidia's AI computing platform) and their intended use in Robotaxi services. However, it focuses on the announcement and plans for deployment rather than any harm or malfunction. There is no indication of injury, rights violations, or operational disruption caused by these AI systems at this stage. The event plausibly could lead to AI-related incidents in the future (e.g., accidents, operational failures), but currently it is a planned deployment and strategic partnership. Hence, it fits the definition of an AI Hazard, as the AI system's use could plausibly lead to harm, but no harm has yet occurred.[AI generated]

The event explicitly mentions the use of generative AI to create sexual deepfake images of real children, which were possessed and shared by the accused. This use of AI directly led to violations of child protection laws and caused harm to the victims and society. Therefore, this qualifies as an AI Incident because the AI system's use directly led to significant harm and legal violations.[AI generated]

The event involves an AI system (Scout) explicitly designed to increase user dependency, which could plausibly lead to harm in the form of mental health issues and addiction. Although no actual injury or harm has been reported, the internal memo's goal to 'make people addicted' and expert concerns about long-term mental health effects indicate a credible risk of future harm. Therefore, this situation qualifies as an AI Hazard rather than an AI Incident, as the harm is potential and not yet realized.[AI generated]

The involvement of AI in cyber attacks is explicitly mentioned, with AI systems assisting threat actors in navigating compromised networks and increasing attack complexity. This use of AI directly leads to harm by enabling more effective and dangerous cyber intrusions, which can disrupt systems and compromise security, fitting the definition of an AI Incident due to harm to property, communities, or critical infrastructure.[AI generated]

An AI system is explicitly involved, as the system uses thermal imaging cameras and AI to monitor for fires and automatically activate fire suppression. The use of this AI system is intended to prevent harm to people, property, and the environment by early fire detection and suppression. Since the system is being introduced to mitigate potential fire hazards and no harm has yet occurred, this qualifies as an AI Hazard, reflecting a plausible future risk being addressed by AI technology.[AI generated]

The attorneys used generative AI to produce legal briefs that contained multiple fabricated cases and misattributed quotations, which are clear inaccuracies caused by AI hallucinations. These false statements were knowingly or recklessly submitted to the court, violating ethical and procedural rules. The AI system's malfunction directly led to a breach of legal obligations and professional conduct, constituting harm under the category of violations of human rights or breach of legal obligations. Hence, this event meets the criteria for an AI Incident.[AI generated]

The event involves an AI system (Google Gemini's voice assistant) whose malfunction or exploitation (prompt injection vulnerability) could directly lead to significant harms including unauthorized physical control of devices, privacy violations, and impersonation attacks. These harms fall under injury or harm to persons (privacy and security breaches) and harm to property or communities (unauthorized control of smart home devices). Since the vulnerability was exploited in demonstrations and poses a direct risk of harm, this qualifies as an AI Incident. The patching and responsible disclosure are complementary information but do not negate the incident classification.[AI generated]