AIM: AI Incidents and Hazards Monitor

This qualifies as an AI Incident because the students’ custom system uses AI facial recognition and language models to directly violate individuals’ privacy rights and extract confidential information, constituting a materialized harm under human rights and privacy categories.[AI generated]

The article explicitly mentions the use of AI systems, including generative AI and AI-generated social bots, to conduct coordinated inauthentic behavior on social media platforms. These AI systems are used to spread disinformation, scams, and manipulate public opinion, which are clear harms to communities and potentially human rights. The involvement of AI in the use phase (operation of bots and content generation) has directly led to these harms. Therefore, this qualifies as an AI Incident under the framework, as the AI systems' use has directly caused significant harm to communities through manipulation and misinformation.[AI generated]

The chatbot’s unauthorized use of the teen’s identity and likeness constitutes misuse of an AI system that directly harmed the family’s emotional well-being and violated the deceased’s personal rights.[AI generated]

The article explicitly involves an AI system (Meta's AI analyzing images and audio from smart glasses) and describes its use of personal data to train AI models. While no direct harm or incident is reported, the extensive data collection and insufficient user awareness create a plausible risk of privacy violations and misuse of personal information. This fits the definition of an AI Hazard, as the development and use of the AI system could plausibly lead to harm related to user privacy and rights. There is no indication of a realized incident or a response update, so it is not an AI Incident or Complementary Information. It is not unrelated because the AI system and its data use are central to the concerns raised.[AI generated]

The UNODC report describes ongoing crimes—fraud, money laundering, forced labor, creation of ultrafake content—explicitly enabled by generative AI systems. These activities have already caused significant financial and human harms. Because AI played a direct and pivotal role in these realized harms, this constitutes an AI Incident.[AI generated]

The article explicitly links AI to the rise and execution of large-scale eCommerce fraud, which causes significant financial harm to merchants and customers. The use of AI-generated deepfakes and synthetic identities to bypass security measures and commit fraud is a direct misuse of AI leading to harm. The harm is materialized and ongoing, not just potential. Although the article also discusses AI-driven fraud detection as a response, the primary focus is on the harm caused by AI-enabled fraud. Hence, this qualifies as an AI Incident under the framework, as AI misuse has directly led to significant harm (financial losses and exploitation).[AI generated]

The event involves the use of an AI system in a critical military defense context, which inherently carries risks of harm if the system malfunctions or is misused. However, the article only reports on successful trials and deployment without any reported injury, violation, or damage. Therefore, it does not meet the criteria for an AI Incident. Instead, it represents a credible potential for harm due to the AI's role in weapon targeting and engagement, qualifying it as an AI Hazard.[AI generated]

The article focuses on the potential future challenges and technical hurdles in scaling AI training infrastructure, including energy grid capacity and data center networking. While it discusses plausible risks and the need for new methods to handle AI training at scale, it does not report any actual harm, malfunction, or violation caused by AI systems. Therefore, it fits the definition of an AI Hazard, as it plausibly could lead to incidents related to energy grid strain or AI system failures in the future, but no incident has yet occurred.[AI generated]

The article describes actual, ongoing AI-driven cyberattacks that have led to data breaches and pose significant threats. The AI systems are being misused to execute these attacks, directly causing harm (violation of privacy, data loss). Thus, it meets the definition of an AI Incident.[AI generated]

The article explicitly mentions the use and planned expansion of AI systems such as facial recognition integrated into surveillance cameras. Although the current system contributes to public safety, the main concern is the plausible future erosion of civil liberties and privacy violations due to intrusive AI surveillance. No direct or indirect harm has been reported as having occurred yet, but the credible risk of such harm is clearly articulated. Hence, the event fits the definition of an AI Hazard, reflecting a credible potential for harm stemming from AI system use in surveillance.[AI generated]

The article describes a research study that assesses the potential impact of AI on different professions, introducing a metric to measure AI's influence on job tasks. It discusses plausible future effects of AI on employment but does not describe any realized harm, malfunction, or misuse of AI systems. Therefore, it fits the definition of an AI Hazard, as it concerns plausible future harm from AI's influence on jobs, without any current incident or realized harm.[AI generated]

An AI system (automated driving) is involved, and the misuse of the system (driver lying down and not monitoring) could plausibly lead to harm such as accidents or injury. However, no actual harm is reported in the event, only a warning against the behavior. Therefore, this qualifies as an AI Hazard due to the plausible risk of harm from misuse of the AI system.[AI generated]

Ecovacs’ devices (which incorporate AI for navigation and mapping) are harvesting intimate home data and using it to train AI without properly informing or securing user consent. This unauthorized data collection and use constitutes a direct harm (privacy and human rights violation) caused by the deployed AI system, fitting the definition of an AI Incident.[AI generated]

The article discusses the intended use and legislative preparation for real-time facial recognition by police, which involves AI systems. However, no actual harm or incident has occurred yet; the event concerns potential future use that could plausibly lead to harms such as privacy violations or rights infringements. Therefore, it qualifies as an AI Hazard rather than an Incident or Complementary Information.[AI generated]

The S-70 is an AI system (autonomous combat UAV) whose malfunction (loss of control/communication) directly led to its destruction. This constitutes an AI Incident since the AI system’s failure caused harm (destruction of property) and operational loss.[AI generated]

The article centers on the announcement and upcoming reveal of Tesla's Robotaxi, an AI-enabled autonomous vehicle system. Although the system is not yet fully autonomous and no harm has been reported, the nature of the technology implies a credible risk of future harm related to autonomous vehicle operation. Therefore, this event qualifies as an AI Hazard due to the plausible future risks associated with the deployment of autonomous driving AI systems. There is no indication of realized harm or incident, nor is the article primarily about responses or updates to past incidents, so it is not an AI Incident or Complementary Information.[AI generated]

Live facial recognition is an AI system under consideration, and although no deployment or harms have yet occurred in Scotland, its use would plausibly lead to violations of privacy, potential civil rights breaches, and misidentification harms (false positives/negatives). The discussion is about preventing future risks rather than reporting a realized incident, making this an AI Hazard.[AI generated]

The article explicitly discusses the development and serial production of KIZILELMA, an unmanned combat aircraft that relies on AI and autonomous technologies for operation. Although no harm or incident is reported, the nature of the system—a weaponized autonomous drone—implies credible potential for harm in the future, such as injury, disruption, or rights violations. The event is not a realized incident but a credible hazard due to the plausible risks associated with AI-enabled autonomous weapons. It is not complementary information because the main focus is on the production start of a potentially harmful AI system, not on responses or updates to past incidents. It is not unrelated because the AI system and its implications are central to the report.[AI generated]

The Super Cruise system is an AI system involved in autonomous driving. The article mentions future plans to upgrade to Level 3 autonomy, which allows drivers to take their eyes off the road, increasing the risk of harm if the system fails or is misused. Since no actual harm or incident is reported yet, but plausible future harm exists due to increased autonomy, this qualifies as an AI Hazard.[AI generated]