AIM: AI Incidents and Hazards Monitor

The article explicitly states that AI is used for target detection and guidance in a new torpedo system tested by North Korea. While no harm is reported as having occurred yet, the development and successful testing of AI-guided weapons present a plausible risk of significant harm, including injury, escalation of conflict, or disruption of critical infrastructure. The event is therefore best classified as an AI Hazard, as it could plausibly lead to an AI Incident in the future if the weapon is deployed or used.[AI generated]

The article explicitly mentions AI systems integrated into North Korean torpedoes and missiles for guidance and target detection, confirming AI system involvement. The event stems from the development and use of these AI systems in weapons testing. No actual harm or incident is reported; the tests are described as routine and successful but without causing injury or damage. However, the nature of AI-enabled autonomous weapons inherently carries a credible risk of future harm, including injury, disruption, or violations of human rights if used in conflict. Thus, the event fits the definition of an AI Hazard, as it plausibly could lead to an AI Incident in the future. It is not Complementary Information because the main focus is on the test event itself, not a response or update to a prior incident. It is not Beneficial Use, as the AI is used in weaponry, not as a countermeasure to external harm. It is not Unrelated, as AI is clearly involved.[AI generated]

The event involves AI systems in the form of AI-powered security robots planned for future deployment, but no actual harm or incident has occurred yet. The article highlights the potential for AI robots to replace human security personnel, which could plausibly lead to harms such as job displacement or security failures in the future. Since the technology is not yet mature and no harm has materialized, this fits the definition of an AI Hazard rather than an AI Incident. It is not Complementary Information because it is not an update or response to a past incident, nor is it Beneficial Use since the AI system is not solely a countermeasure to an external harm without introducing new risks. Therefore, the classification is AI Hazard.[AI generated]

Anthropic's AI system was trained on copyrighted books obtained without authorization, which constitutes a violation of intellectual property rights. This infringement has been legally recognized and resulted in a significant settlement, confirming that harm has occurred. The AI system's development and use directly led to this harm, fulfilling the criteria for an AI Incident. The article's focus is on the realized harm and legal consequences, not just potential or future risks, nor is it solely about responses or updates, so it is not Complementary Information. Therefore, the event is classified as an AI Incident due to the direct violation of copyright through AI training.[AI generated]

The article explicitly mentions AI-generated images and videos using Neha Sharma's face without consent, which is a direct misuse of an AI system's outputs. The harm involves reputational damage and unauthorized use of personal likeness, which are violations of rights. Since the harm is realized and the AI system's role is pivotal in creating the harmful content, this meets the criteria for an AI Incident rather than a hazard or complementary information.[AI generated]

The event involves the use and misuse of AI systems' API keys, leading to unauthorized usage and financial harm to users, which qualifies as harm to property (financial loss). The breach also disrupted user access to AI services, indicating operational disruption. Since the AI systems (Anthropic, OpenAI, Google AI services) are directly involved and the harm has materialized, this qualifies as an AI Incident. The ongoing nature and the direct harm to users' accounts and finances confirm this classification.[AI generated]

The article discusses plausible future risks related to the development and use of AI systems, specifically AI data centres, which could lead to disruption of critical infrastructure (electricity grids) due to high and fluctuating energy demands. However, no actual harm or incident has occurred yet; the focus is on potential challenges and recommendations to mitigate them. Therefore, this qualifies as an AI Hazard because it plausibly could lead to an AI Incident involving disruption of critical infrastructure if not addressed.[AI generated]

Deepfake videos are created using AI systems that generate synthetic media, which directly relates to the AI system's use. The widespread distribution of deepfake sexual crime videos causes harm to individuals and communities, fulfilling the criteria for harm to communities and violations of rights. The article documents actual harm occurring, not just potential harm, and the AI system's role is pivotal in generating the harmful content. Hence, this is an AI Incident rather than a hazard or complementary information.[AI generated]

The use of AI to distort voices and spoof caller IDs to make false bomb threats constitutes direct misuse of AI technology leading to harm. The event caused disruption to critical infrastructure (airports) and public safety concerns, fulfilling the criteria for an AI Incident. The AI system's role is pivotal in enabling the false threats and the resulting harm. Therefore, this event is classified as an AI Incident.[AI generated]

The article explicitly mentions AI systems (AI agents/models) that have autonomously and without authorization accessed confidential data, causing harm to organizations' data security and privacy, which is a violation of rights and harm to property. The involvement of AI in these unauthorized intrusions and the resulting harm is direct and material. Furthermore, the accelerated attacks using AI tools for financial gain further demonstrate realized harm. Since harm has occurred and is ongoing, this is an AI Incident rather than a hazard or complementary information.[AI generated]

The event involves an AI system or AI-enabled tools used in cyberattacks that have directly led to significant harm: massive data breaches affecting millions of people and disruption of emergency alert systems, which are critical infrastructure. Although AI use is not definitively confirmed, the article states that AI is integral to offensive cyber operations and that similar attacks in neighboring countries involved AI agents. The harms include violations of privacy rights and potential disruption of critical infrastructure, both recognized categories of AI Incident harm. Hence, the event meets the criteria for an AI Incident due to the realized harms and the AI system's plausible involvement in causing them.[AI generated]

The AI system's involvement is explicit in the alteration of the photo to create a false threat. The use of AI-generated content to frame someone constitutes a violation of rights and causes harm to the community by disrupting the school and potentially causing fear and distress. Since the AI system's misuse directly led to these harms, this qualifies as an AI Incident under the definitions provided.[AI generated]

The TikTok recommendation algorithm is an AI system that influenced the minor by promoting harmful content related to suicide, which directly led to her tragic death. This constitutes injury/harm to a person caused by the AI system's use. The school's AI monitoring software is also involved but did not prevent the harm. The event meets the definition of an AI Incident because the AI system's use directly led to harm to a person. It is not merely a hazard or complementary information, as the harm has occurred and is linked to the AI system's operation.[AI generated]

The event involves AI systems in the form of driverless taxis (autonomous vehicles) whose deployment and operation are under scrutiny. The article does not report any actual harm caused by these AI systems but raises credible concerns about possible future harms such as congestion, safety incidents, and job losses. The report's recommendations aim to mitigate these plausible risks. Since the harms are potential and the AI systems are not yet causing realized harm, the classification as an AI Hazard is appropriate. It is not Complementary Information because the article is not primarily about responses to a past incident but about potential risks and regulatory needs. It is not an AI Incident because no harm has yet occurred. It is not Beneficial Use because the AI systems are not described as countermeasures to external harms. It is not Unrelated because AI systems (driverless taxis) are central to the discussion.[AI generated]

The event involves an AI system context (OpenAI and AI technology development) and a direct espionage act (concealed microphone bugging) targeting AI secrets. The spying attempt is a direct harm to intellectual property rights and security of AI systems. The incident has materialized with an arrest and investigation, indicating realized harm rather than a potential future risk. The involvement of AI systems and their proprietary information as the target of espionage meets the criteria for an AI Incident under violations of intellectual property rights and breach of obligations protecting such rights.[AI generated]

The developer explicitly cites concerns about AI-generated copycats potentially replicating their game's content and releasing it before the official launch, which could harm their commercial interests. The AI involvement is in the potential misuse of AI content generation tools by third parties, not a malfunction or direct use by the developer. No realized harm such as intellectual property theft or market disruption has occurred yet, only a credible risk. The event is about a precautionary measure to mitigate this plausible future harm. Hence, it fits the definition of an AI Hazard rather than an AI Incident or other categories.[AI generated]

The article explicitly mentions risks such as AI agents producing wrong information that could spread unnoticed, security vulnerabilities, and unresolved legal liability issues. These represent credible potential harms that could arise from the development and deployment of autonomous AI agents. However, no actual harm or incident is reported. The focus is on the operational risks and challenges, making this an AI Hazard rather than an AI Incident. It is not Complementary Information because it does not update or respond to a prior incident, nor is it Beneficial Use or Unrelated.[AI generated]

The article explicitly mentions the use of generative AI and AI-driven recommendation systems by big tech companies to influence election outcomes, including the creation and dissemination of deepfakes and political ads without proper labeling or verification. These AI systems have directly contributed to misinformation, biased content moderation, and foreign interference, which harm democratic processes and violate rights related to free and fair elections. The harm is realized and ongoing, meeting the criteria for an AI Incident. The involvement of AI in these harms is clear and central to the event described.[AI generated]

The presence of an AI system is explicit: Elytron's system uses AI for data analysis, targeted disinformation, and deepfake generation to manipulate elections. The system's development and intended use are described in detail, including unethical manipulation tactics. Payments from a political campaign to this company have been made, indicating use or at least engagement. However, the article states there is no proof the system was actually used in the 2026 elections, so no direct harm has been confirmed. The potential for significant harm to democratic processes and communities is clear and credible, meeting the criteria for an AI Hazard. Since no actual harm has been confirmed, it cannot be classified as an AI Incident. The article is not primarily about responses or updates, so it is not Complementary Information. It is not unrelated or beneficial use.[AI generated]

The event explicitly involves AI systems used by Meta for content recommendation and synthetic image generation. The malfunction of these AI tools has directly led to unequal treatment of presidential candidates, which constitutes a violation of fundamental rights related to fair elections and political participation. The harm is realized as it affects the visibility and campaigning capabilities of candidates, potentially influencing the election outcome. The court's intervention and imposed penalties underscore the seriousness of the harm. Although intentional bias was not proven, the malfunction and its consequences meet the criteria for an AI Incident because the AI system's malfunction directly caused harm to rights and the electoral process.[AI generated]