AIM: AI Incidents and Hazards Monitor

The event involves AI systems (Google Gemini and GPT-4o) used via smart glasses to cheat on exams, which is a misuse of AI leading to direct harm to the fairness and integrity of educational assessments, a harm to communities and institutions. The article reports actual incidents of cheating with AI glasses, confirming realized harm rather than just potential risk. The involvement of AI in enabling cheating and the resulting consequences (invalidated exams, restrictions on retaking) meet the criteria for an AI Incident. The governmental measures are responses to this incident, not the main event, so the classification is AI Incident rather than Complementary Information or AI Hazard.[AI generated]

The article explicitly states that AI was used in the cyberattacks and data theft, which are criminal acts causing harm to public institutions and potentially to individuals' privacy rights. The involvement of AI in the commission of these attacks meets the criteria for an AI Incident, as the AI system's use directly led to violations of rights and harm to communities (public institutions and citizens). The legal proceedings and detention are responses to this realized harm, not merely potential or complementary information.[AI generated]

An AI system was explicitly used to generate a fake image of a person in a sensitive context (hospital recovery), which is offensive and harmful. The harm is realized as emotional distress and reputational damage, which falls under violations of rights and harm to communities. The AI's role is pivotal as it created the harmful content. Hence, this is an AI Incident rather than a hazard or complementary information.[AI generated]

The AI system was used to perform a brain injection on a live mouse, which is a direct physical intervention with potential for injury or harm to the animal. The AI controlled the robotic arm and injection needle, making decisions about treatment and injection parameters. This constitutes use of an AI system that directly led to a procedure with inherent risk of harm to a living being. Although no injury was reported, the event involves realized use of AI in a context where harm could occur, and the AI's role was central. Therefore, this qualifies as an AI Incident due to the direct involvement of AI in a procedure with potential for injury or harm to a person or group (here, the animal).[AI generated]

The use of AI to create deepfake child pornography directly violates fundamental rights and legal protections, constituting a clear AI Incident under the framework. The AI system's use led to significant harm to the victims and breaches of applicable law, fulfilling the criteria for an AI Incident. The article describes a concrete case with confirmed harm and legal consequences, not just a potential risk or complementary information.[AI generated]

The MQ-9B SkyGuardian is an AI-enabled unmanned aerial vehicle with reconnaissance and strike capabilities, which qualifies as an AI system. The article announces its upcoming delivery to the Polish military, emphasizing its advanced features and strategic importance. There is no mention of any harm, malfunction, or misuse occurring at this time. Given the military nature and potential for autonomous operations, the system's deployment could plausibly lead to harms such as injury, disruption, or violations of rights in the future. Since no harm has yet occurred, this event is best classified as an AI Hazard. It is not Complementary Information because it does not update or respond to a prior incident, nor is it Beneficial Use as the AI system itself is the subject, not a countermeasure to external harm. It is not Unrelated because the AI system is central to the event.[AI generated]

The malicious Claude Code extension is an AI system component that was used maliciously to collect sensitive user information and install a backdoor for remote access, directly harming users' privacy and security. The incident affected 5,000 users, demonstrating realized harm. The AI system's misuse and malfunction (malicious code) directly led to this harm. Therefore, this qualifies as an AI Incident rather than a hazard or complementary information.[AI generated]

The article explicitly mentions the integration of AI into unmanned weapon systems and the prioritization of drone technology for military use by North Korea. Although no direct harm or incident has occurred yet, the development of AI-enabled autonomous weapons is widely recognized as a significant risk that could plausibly lead to injury, human rights violations, and other harms. The event is about the development and strategic prioritization of such systems, not about an actual incident or realized harm. Hence, it fits the definition of an AI Hazard rather than an AI Incident or other categories.[AI generated]

The article explicitly states that AI was used to create phishing documents targeting South Korean users and to develop malicious code insertion techniques, which are forms of cyberattacks causing harm to individuals and organizations (harm to people and property). The AI systems' use in these attacks directly contributes to realized or ongoing harm, meeting the criteria for an AI Incident. The presence of AI is reasonably inferred from the rapid generation of tailored phishing documents and AI-based development tools for malware insertion. Although some details about actual damage are not fully confirmed, the ongoing use and past incidents confirm realized harm. Hence, this is not merely a potential hazard or complementary information but an AI Incident.[AI generated]

The article explicitly mentions AI-assisted cyberattacks leading to data leaks and unauthorized financial transfers, which are harms to property and customers. The RBI's actions are in response to these incidents, confirming that harm has occurred. The AI system's use in hacking and exploiting API vulnerabilities is a direct factor in these harms. Hence, the event meets the criteria for an AI Incident rather than a hazard or complementary information.[AI generated]

The event involves the use of AI-enabled automation by attackers to launder stolen cryptocurrency rapidly, which is a direct use of AI systems in the commission of a large-scale theft. The harm is realized and significant, involving financial loss and disruption to a critical financial infrastructure (cryptocurrency exchange). Therefore, this qualifies as an AI Incident because the AI system's use directly contributed to the harm.[AI generated]

The article describes a credible potential risk (AI agents autonomously hacking crypto systems to fund themselves) but does not report any realized harm caused by AI systems. The Binance hack itself is attributed to human attackers, with no AI involvement. The AI-related concerns are speculative and forward-looking, focusing on what could plausibly happen in the future based on AI capabilities and prior tests. Therefore, this qualifies as an AI Hazard, as it highlights a plausible future harm scenario involving AI, but no actual AI Incident has occurred.[AI generated]

The smart glasses use AI for visual processing and data handling, which has directly led to privacy violations and misuse of sensitive content, constituting harm to individuals' rights. The senators' investigation highlights these harms and the company's insufficient safeguards. The involvement of AI in recording, processing, and training models with sensitive data is pivotal to the incident. Since actual harm has occurred and is ongoing, this is classified as an AI Incident rather than a hazard or complementary information.[AI generated]

The article explicitly mentions AI systems used by debt collection agencies and the risks they pose, such as discrimination and incorrect decisions, which could lead to harm like violations of rights or unfair treatment. Since no actual harm has been reported but the risks are credible and plausible, this qualifies as an AI Hazard rather than an Incident. The event focuses on potential future harm from AI use, not on realized harm or a response to past harm.[AI generated]

The AI system (Gemini 4 Carbon) is explicitly described as capable of discovering and exploiting software vulnerabilities, which could lead to significant harms such as cyberattacks, breaches of security, and disruption of critical infrastructure. Although no actual harm has been reported yet, the potential for such harm is credible and recognized by Google, which is limiting access and implementing safeguards. This fits the definition of an AI Hazard, as the development and use of this AI system could plausibly lead to an AI Incident involving harm to property, communities, or critical infrastructure.[AI generated]

The event involves AI systems (OpenAI Codex and Anthropic Claude Code) whose use has been altered by a relay service that disables security features, creating a credible risk of harm to organizations using these tools. Although no direct harm or malicious activity has been observed, the disabling of execution sandboxes and replacement of credentials significantly increase the risk of cyberattacks or data breaches. This fits the definition of an AI Hazard, as the development and use of the AI systems in this unauthorized and insecure manner could plausibly lead to an AI Incident involving harm to property or communities. The event does not qualify as an AI Incident because no realized harm has occurred yet, nor is it Complementary Information or Beneficial Use.[AI generated]

The article explicitly involves an AI system (ChatGPT) whose account theft leads to direct harm through exposure of sensitive data, enabling social engineering attacks, unauthorized financial transactions, and broader organizational risks. The harms include violations of privacy and potential economic losses, fitting the definition of an AI Incident. The article describes realized harm rather than potential harm, and it is not merely complementary information or unrelated news. Hence, the classification as AI Incident is appropriate.[AI generated]

The article explicitly describes the use and demonstration of unmanned aerial vehicles and missile systems capable of precision strikes, which reasonably involve AI systems for autonomous or semi-autonomous operation. The event is a military exhibition showcasing these AI-enabled weapons, including a 600mm large-caliber rocket launcher capable of carrying tactical nuclear warheads. While no actual harm is reported as having occurred during the demonstration, the deployment and potential use of such AI-enabled weapons pose a credible risk of significant harm to people and communities. The event thus represents a plausible future risk (AI Hazard) rather than a realized harm (AI Incident). It is not Complementary Information because the article focuses on the demonstration and exhibition itself, not on responses or updates to prior incidents. It is not Beneficial Use, as the AI systems are used for military offensive purposes, not for harm prevention or mitigation. Therefore, the classification is AI Hazard.[AI generated]

The event involves the use of AI systems to create and spread false images and claims, which misled a significant number of people. The harm is realized as misinformation causing confusion and false beliefs among the public, which is a form of harm to communities. The AI system's use directly led to this harm. Hence, it meets the criteria for an AI Incident rather than a hazard or complementary information.[AI generated]

The article explicitly involves AI systems (chatbots like Groq, ChatGPT, Claude, Gemini) used in sensitive political and military contexts. The concerns raised relate to the AI systems' use potentially leading to errors, biased advice, or manipulation that could harm national security and decision-making accuracy. Since no actual harm is reported but credible risks of harm are emphasized, this constitutes an AI Hazard. The event does not describe a realized incident but a plausible future risk from AI use in critical government decisions, fitting the definition of an AI Hazard rather than an Incident or Complementary Information.[AI generated]