AIM: AI Incidents and Hazards Monitor

The event involves the use of AI systems to generate legal research that included fabricated citations, which is a misuse of AI outputs leading to a breach of legal and professional obligations. The direct consequence was a severe professional penalty, indicating realized harm. The AI system's role is pivotal as the fabricated citations originated from AI-generated content. This meets the criteria for an AI Incident due to the direct link between AI misuse and harm to legal rights and professional standards.[AI generated]

The event involves the use of an AI system (autonomous driving technology) in real-world testing, which inherently carries risks of harm such as accidents or safety issues. Since the article does not report any actual harm or incidents but discusses the pilot testing phase and the intention to validate safety and compliance, it fits the definition of an AI Hazard. The potential for future harm is credible given the nature of autonomous vehicle deployment in traffic, but no direct or indirect harm has yet materialized. Hence, it is not an AI Incident. It is also not Complementary Information because it is not an update or response to a prior incident, nor is it unrelated as it clearly involves AI systems.[AI generated]

The event involves AI systems (high-end AI chips and servers) and their export control, which is a governance and policy response to potential risks associated with AI technology proliferation. Since no actual harm or incident has occurred, but there is a plausible risk that unauthorized export could lead to misuse or harm, this qualifies as an AI Hazard. The article primarily reports on the potential for harm and the regulatory measures being considered, rather than an incident or complementary information about past events or responses.[AI generated]

The Saronic Corsair is explicitly described as an AI-powered autonomous surface vessel used in a real-world rescue operation. The AI system's involvement in the rescue directly relates to harm prevention (injury or death of the helicopter crew). Since the AI system's use directly led to the rescue and thus prevented harm, this qualifies as an AI Incident under the definition of an event where AI use has directly or indirectly led to injury or harm (in this case, harm was averted but the event involved real risk and AI's pivotal role). The article does not describe a hazard or potential future harm but an actual event involving AI use with direct impact on human safety.[AI generated]

The event involves AI systems explicitly (AI-based autonomous combat drones, AI-enabled maintenance systems) and their development and intended use in military applications. While no harm has occurred, the deployment and demonstration of AI-enabled unmanned combat systems and integrated manned-unmanned air combat systems plausibly could lead to harms such as injury, disruption, or rights violations in future conflict scenarios. The article does not report any realized harm or incident, so it is not an AI Incident. It is not merely complementary information because the focus is on unveiling new AI military technologies with potential risks. Hence, it fits the definition of an AI Hazard.[AI generated]

The article explicitly mentions the use of AI for risk analysis and tailored intervention in juvenile offender management, indicating the involvement of an AI system in the development and use phases. However, the AI system is still in the development or early deployment phase, with no reported incidents of harm or rights violations. The focus is on strengthening infrastructure and preventive measures using AI, which could plausibly lead to harm if misused or malfunctioning but currently serves as a preventive tool. Hence, this qualifies as an AI Hazard rather than an AI Incident or Complementary Information. It is not unrelated because AI involvement is clear and central to the new policy.[AI generated]

Meta's content moderation relies heavily on AI systems to detect and remove harmful content. The rollback of these AI-driven moderation policies led to a measurable increase in violent threats and harassment, which are direct harms to individuals and communities. The report from CCDH provides evidence of this causal link. The AI system's reduced enforcement capability (whether by design or malfunction) directly contributed to the increase in harmful content, fulfilling the criteria for an AI Incident. The event is not merely a potential risk or a complementary update but a realized harm linked to AI system use.[AI generated]

The article explicitly mentions AI-generated fake adverts that impersonate public figures and central banks, which are being used to mislead and potentially scam users. The AI system's outputs (fake videos) have directly caused harm by deceiving the public and enabling fraudulent schemes. This fits the definition of an AI Incident because the AI system's use has directly led to harm to communities and violations of rights through digital deception and fraud. The involvement of AI in generating the fake content and the resulting harm is clear and ongoing, not merely a potential risk or complementary information.[AI generated]

The AI system is explicitly involved as the cheating tool (AI glasses with AI functions). The event reports actual use of these AI systems to gain unfair advantage in exams, which is a violation of rules and harms the integrity of the examination process, thus constituting harm to communities and institutions. The harm is realized, not just potential. Hence, this is an AI Incident.[AI generated]

The event involves an AI system explicitly described as analyzing transaction sequences to detect and predict financial fraud, which is a direct use of AI in preventing harm. The AI system's deployment and its impact on reducing fraud cases indicate its role in harm management. Although the harm is prevented rather than caused, the AI system's involvement in the context of harm (financial fraud) is central. According to the definitions, AI Incidents include events where AI systems' use directly or indirectly leads to harm or its prevention in critical areas such as financial fraud. Hence, this qualifies as an AI Incident rather than a hazard or complementary information.[AI generated]

The article centers on the risks and challenges posed by AI-generated disinformation to democracy and electoral freedom, which could plausibly lead to harm such as violations of rights and harm to communities. However, it does not describe a specific realized harm or incident caused by AI, but rather a warning and institutional efforts to address the issue. Therefore, it fits the definition of an AI Hazard, as it plausibly could lead to harm but no concrete incident is detailed.[AI generated]

The article discusses the planned deployment of an AI system (AI electronic fences with image recognition) for school safety monitoring. Since the AI system is not yet in use and no harm or incident has occurred, but the system could plausibly lead to prevention of harm or, conversely, if malfunctioning, could lead to safety issues, this qualifies as an AI Hazard. The event is about a proposed policy and planned AI use with potential future impact, not an incident or complementary information about an existing incident.[AI generated]

The article clearly describes AI systems (autonomous drones, algorithmic targeting) being used in active conflicts, resulting in increased deaths and humanitarian crises. The AI systems' involvement in lethal decision-making with reduced human oversight has directly led to harm to people and communities. This meets the definition of an AI Incident because the AI system's use has directly caused injury and harm to groups of people. The report also discusses the broader geopolitical and economic context but the central AI-related harm is the lethal autonomous weapons and AI-driven targeting causing real-world casualties.[AI generated]

The video is an AI-generated manipulated content (deepfake) falsely portraying military leaders, which is a misuse of AI technology. Although the video is circulating, the military denies any official communication and warns against misinformation. There is concern about potential harm to public confidence and national security, but no direct or indirect harm has occurred yet. Therefore, this event represents a plausible future risk of harm due to AI misuse, qualifying it as an AI Hazard rather than an AI Incident. It is not merely complementary information because the main focus is on the fabricated AI content and its potential to cause harm, not on responses or updates to a past incident.[AI generated]

The LITAVR drone incorporates an AI system for automatic target identification and interception, which is explicitly described as automating up to 95% of the interception process. The system's use has directly led to the destruction of enemy drones, which is harm to property and potentially to communities by defending against attacks. The AI system's development, deployment, and use in combat have caused realized harm, meeting the criteria for an AI Incident. The event is not merely a potential hazard or complementary information but a report of an AI system actively causing harm in a military context.[AI generated]

The article clearly involves AI systems used in medical diagnosis and treatment recommendation, which can directly influence patient health outcomes. Although no specific AI-caused harm has yet been reported, the scenarios described (e.g., missed tumor, incorrect medication dosage) represent credible risks of injury or death to patients, fitting the definition of an AI Hazard. The focus is on the plausible future harms and legal liability concerns arising from AI use in healthcare, rather than on a realized incident or a response to one. Therefore, this event is best classified as an AI Hazard.[AI generated]

An AI system is explicitly involved, used for monitoring and identifying fake medical practitioners, which directly relates to health and safety. The system's role is to generate alerts that help prevent harm to public health by detecting irregularities early. Since no harm has yet occurred but the system is designed to prevent potential harm, this qualifies as an AI Hazard. The event does not describe any realized harm or incident but focuses on the potential to prevent harm through AI use.[AI generated]

The HALCON software is an AI system used for image inspection. The event involves the unauthorized use and alleged cracking of this AI software, leading to a violation of intellectual property rights. This harm has already materialized, as evidenced by criminal prosecution and civil lawsuits. The courts have confirmed the infringement and rejected the defendant's claims. Hence, the event meets the criteria for an AI Incident due to realized harm (violation of intellectual property rights) caused by the use and misuse of an AI system.[AI generated]

The event involves the use of an AI system (autonomous drone interception technology) that has been deployed and tested in real combat conditions. The AI system's autonomous operation directly contributes to the destruction of enemy drones, which is a form of military action with potential physical harm implications. Since the AI system is actively used in combat to destroy targets, this constitutes an AI Incident due to the direct involvement of AI in causing harm (destruction of enemy drones) in a real-world scenario.[AI generated]

An AI system (the drowning detection system) was explicitly mentioned and was in use at the time of the incident. The system's failure to issue an alert when the drowning occurred constitutes a malfunction. This malfunction indirectly contributed to the harm (the death of the swimmer) because the system did not assist in timely rescue. The event involves direct harm to a person caused in part by the AI system's malfunction, meeting the criteria for an AI Incident.[AI generated]