AIM: AI Incidents and Hazards Monitor

The AI system Mythos was used in a controlled penetration test and successfully exploited vulnerabilities rapidly, demonstrating a capability that could plausibly lead to significant cybersecurity incidents if misused. No actual unauthorized harm or breach occurred, so it is not an AI Incident. However, the demonstrated capability and the government's restrictive response indicate a credible risk of future AI-enabled cyberattacks, fitting the definition of an AI Hazard. The article does not focus on remediation or governance responses alone, so it is not Complementary Information. The event is clearly related to an AI system and its potential for harm, so it is not Unrelated.[AI generated]

The drones are AI systems as they are described as autonomous defense systems piloted by AI. The procurement and deployment of such a large number of AI-enabled combat drones represent a credible risk of harm, given their military use and potential for lethal applications. Although no harm has yet occurred, the nature of these AI systems and their intended use in combat plausibly could lead to significant harms such as injury, disruption, or violations of rights. Therefore, this event qualifies as an AI Hazard due to the plausible future harm from the deployment of AI-powered combat drones.[AI generated]

The article explicitly mentions the use of algorithms that create a "rabbit-hole effect" to keep children engaged, which implies AI systems are involved in content recommendation and user engagement. The investigation is regulatory and concerns potential breaches of child safety and well-being, indicating plausible future harm from AI system use. Since no confirmed harm or incident has been established yet, and the event centers on the investigation and regulatory process, this qualifies as an AI Hazard rather than an AI Incident. It is not merely complementary information because the investigation itself highlights a credible risk of harm from AI systems.[AI generated]

The event involves an AI system (Claude chatbot) experiencing malfunction leading to service disruption and degraded functionality. While no direct physical harm or legal violation is reported, the malfunction causes significant disruption to users relying on the AI system for various tasks. This fits the definition of an AI Incident as the AI system's malfunction is directly leading to harm in the form of service disruption and operational impact to users. The harm is realized and ongoing, not merely potential. Therefore, this event qualifies as an AI Incident.[AI generated]

The article explicitly mentions an AI system (Safe Pro's AI-powered threat detection technology) being integrated into autonomous UGVs for real-time detection of explosive threats. The AI system is under development and planned deployment, with no reported incidents or harms resulting from its use so far. The intended use in military autonomous vehicles for detecting landmines and explosives implies a credible potential for harm if the AI system malfunctions or is misused, such as failure to detect threats or false positives leading to operational issues. However, since no actual harm or incident is reported, the event is best classified as an AI Hazard, reflecting the plausible future risk associated with the AI system's deployment in a high-stakes environment.[AI generated]

The article explicitly discusses AI systems in the form of autonomous drones used for military and defense purposes, including autonomous mission execution and counter-UAS operations. However, it does not describe any realized harm, injury, rights violations, or disruptions caused by these AI systems. The content centers on the market growth, technological capabilities, and strategic deployment of these systems, which implies a credible risk of future harm due to their autonomous military use. Therefore, the event fits the definition of an AI Hazard, as the development and proliferation of autonomous military drones could plausibly lead to AI Incidents in the future, but no direct or indirect harm has yet occurred as per the article.[AI generated]

The article explicitly mentions the use of generative AI to create synthetic videos that sexualize and stereotype women by nationality, which is a form of gender-based disinformation and objectification. This use of AI has directly led to harm by reinforcing harmful societal stereotypes and contributing to misogyny and unrealistic beauty standards, which affect communities and violate rights related to gender dignity. The AI system's development and use are central to the creation and spread of this harmful content. Hence, the event meets the criteria for an AI Incident as it involves realized harm caused directly by AI-generated content.[AI generated]

The article explicitly involves AI systems (generative AI chatbots) used in healthcare. It details how their use has led or could lead to harms including health risks from incorrect advice and privacy violations from data exposure. The discussion of actual cases where models revealed user data and the warning about cyberattacks indicate realized or ongoing harms. Therefore, this qualifies as an AI Incident because the AI system's use has directly or indirectly led to harms to health and violations of privacy rights. The article is not merely a warning about potential future harm (AI Hazard) nor a general update or response (Complementary Information).[AI generated]

The Tesla FSD system is an AI system as it uses deep learning for driver assistance and autonomous control functions. The article centers on Traficom's safety and regulatory assessment of the system's conditional approval, highlighting potential risks such as driver inattentiveness, system errors, and speeding enabled by the AI system. However, no actual incidents or harms have been reported; the system is under evaluation before full approval. This fits the definition of an AI Hazard, where the AI system's use could plausibly lead to harm (e.g., accidents due to system errors or driver misuse), but no direct or indirect harm has yet occurred. The article does not describe any realized injury, rights violation, or property/environmental harm caused by the AI system. It is not merely complementary information because the focus is on the potential safety risks and regulatory decision-making regarding the AI system's deployment, not on responses to past incidents or general AI ecosystem updates.[AI generated]

The event involves the use and manipulation of AI systems (search algorithms and AI language models) through the creation of fake content designed to mislead and propagate propaganda. The harm is realized as misinformation is actively spread and AI systems are trained on biased data, affecting the information environment and communities. This fits the definition of an AI Incident because the AI systems' development and use have directly led to harm to communities through misinformation and propaganda dissemination. The involvement of AI in the manipulation and amplification of false content is explicit and central to the harm described.[AI generated]

The article explicitly mentions AI systems being used by companies to process personal data and train models without adequate transparency, leading to complaints and regulatory investigations. This indicates that the AI systems' use has directly led to violations of data protection laws and rights, fulfilling the criteria for an AI Incident. The harm is realized in the form of rights violations and legal breaches, not merely potential or hypothetical, and the event involves the AI systems' use rather than just their development or potential misuse.[AI generated]

The event involves the use of an AI system (Kalibrate's AI-based pricing tool) in a way that allegedly leads to harm by artificially inflating gas prices, which is a violation of antitrust law and harms consumers financially. The AI system's use is central to the alleged harm, making this an AI Incident under the framework because the AI's use directly led to a breach of legal obligations and economic harm to a community of people.[AI generated]

The event involves AI systems implicitly through the mention of drone and robotics companies, which typically use AI for autonomous functions. The export control is a preventive measure to avoid potential harms such as military misuse or proliferation risks, which could lead to harm to national security or other harms if these AI systems were misused. Since no actual harm has occurred yet, and the event is about regulatory action to prevent such harm, it fits the definition of an AI Hazard. It is not Complementary Information because it is not an update or response to a past incident, nor is it unrelated as it clearly involves AI-related technologies and their control.[AI generated]

The article focuses on a government draft regulation planning to embed AI in public programs, outlining intended uses and potential benefits, as well as acknowledging possible AI-related risks. There is no report of actual harm or malfunction caused by AI systems at this stage. The event is prospective and concerns plausible future impacts of AI adoption, making it an AI Hazard rather than an Incident. It is more than general AI news because it details a specific government plan with potential for future harm or benefit, but no harm has yet occurred.[AI generated]

The event involves AI systems explicitly in the creation of fraudulent applications that cause harm by deceiving users and stealing information, fulfilling the criteria for an AI Incident. The harm is realized (financial losses, privacy violations), and the AI's role is pivotal in enabling sophisticated fraud that is harder to detect. Therefore, this is not merely a potential risk or complementary information but a direct AI Incident.[AI generated]

The event involves the development and intended use of an AI system for traffic light control, which could plausibly lead to future harms or benefits. However, there is no indication that any harm or malfunction has occurred so far. The article focuses on the planned implementation and expected positive outcomes rather than any realized harm or incident. Therefore, this qualifies as an AI Hazard because the AI system's deployment could plausibly lead to incidents (e.g., traffic accidents, mismanagement) in the future, but no incident has yet occurred.[AI generated]

The event involves the use of an AI system (ChatGPT) to generate fake exam papers, which were then sold to mislead and cheat buyers. This use of AI directly caused harm by facilitating fraud and deception, which falls under violations of legal obligations and harms to communities (misleading students and undermining examination integrity). Therefore, this qualifies as an AI Incident because the AI system's use directly led to realized harm through fraudulent activity.[AI generated]

An AI system is involved as the tracking program collects detailed user interaction data to train AI models. The use of this AI system (data collection and training) has directly led to harm in the form of privacy violations and employee distress, which can be considered a violation of rights and harm to individuals. The potential exposure of data accessible to anyone inside the company further supports the presence of harm. Therefore, this qualifies as an AI Incident due to realized harm linked to the AI system's use.[AI generated]

The event explicitly involves an AI system (deepfake technology) used to generate manipulated images that mislead the public. The use of this AI-generated content has directly led to the spread of disinformation with potential harm to the electoral process and democratic debate, which qualifies as harm to communities and violation of rights. Therefore, this constitutes an AI Incident because the AI system's use has directly led to harm through disinformation and manipulation of political information.[AI generated]

The event involves the use of AI-enabled robotic systems for delivery, which qualifies as an AI system. The replacement of human workers by robots could plausibly lead to significant harm in terms of employment loss and social disruption, which falls under harm to communities. Since the harm is anticipated but not yet realized, and the company is actively preparing for this transition, the event is best classified as an AI Hazard. There is no indication of an actual incident or realized harm, nor is the article primarily about responses or governance measures, so it is not Complementary Information.[AI generated]