AIM: AI Incidents and Hazards Monitor

The event involves AI systems (generative AI creating deepfake videos and language models) whose use has directly led to harm to communities by spreading false information that threatens election integrity and democracy. The harm is realized and ongoing, not merely potential. The article explicitly states that AI-generated videos are the main vector of disinformation, which is a clear harm to communities and democratic institutions. Therefore, this is an AI Incident rather than a hazard or complementary information.[AI generated]

The article involves AI systems (AI shopping bots) whose use could plausibly lead to harm such as financial fraud or scams, as highlighted by the banks. Since no actual harm or incident has been reported, but credible risks are identified, this qualifies as an AI Hazard. The banks' planned discussions with policymakers and calls for transparency and security measures are responses to these potential risks but do not themselves constitute an incident or complementary information about a past incident.[AI generated]

An AI system was explicitly used to alter the photograph, changing racial appearances and misrepresenting individuals. This alteration caused harm by erasing the identity of a student and misleading the public about the university's diversity, which can be seen as harm to communities and a violation of rights. The harm has materialized as the affected student expressed upset and the university is responding to the incident. The AI system's misuse directly led to this harm, fulfilling the criteria for an AI Incident.[AI generated]

The article centers on warnings and concerns about potential harms AI could cause to children, reflecting a credible risk but no specific event of harm has been reported. The involvement of AI systems (chatbots and AI search features) is explicit, and the potential harms are serious and plausible. Since no actual harm has occurred yet, and the focus is on the risk and calls for safety standards and regulation, this qualifies as an AI Hazard. It is not Complementary Information because the main narrative is not about responses to a past incident but about the emerging risk and need for action. It is not an AI Incident because no direct or indirect harm has been reported as having occurred.[AI generated]

The event involves an AI system (EvilTokens) used maliciously to conduct phishing attacks that have directly caused harm to individuals and organizations through fraud and unauthorized access to email accounts. The AI's role in analyzing inboxes and automating attack strategies is central to the harm caused. Therefore, this qualifies as an AI Incident due to realized harm stemming from the AI system's use in cybercrime.[AI generated]

The article explicitly describes an AI system generating news content and avatars, which is a clear AI system involvement. The concerns raised by journalists and media organizations about job losses and content appropriation indicate potential harms related to labor rights and intellectual property rights. However, the article does not document actual harm or incidents such as layoffs or legal breaches that have already occurred. Instead, it focuses on the plausible future risks and ethical challenges posed by the AI system's deployment. According to the definitions, this situation qualifies as an AI Hazard because the AI system's use could plausibly lead to significant harms, but these harms have not yet materialized. The event is not Complementary Information because the main focus is not on responses or updates to past incidents but on the emergence of a new AI-driven media outlet and the concerns it raises. It is not Beneficial Use because the AI is not deployed as a countermeasure to external harm but as a primary content generator, which may itself cause harm. Therefore, the correct classification is AI Hazard.[AI generated]

The event clearly involves an AI system used to create deepfake images, which is a form of AI-generated content manipulation. The harm is direct and significant, including violations of the victim's rights, psychological harm, and reputational damage. The creation and dissemination of non-consensual deepfake pornography constitute a violation of fundamental rights and cause harm to the individual and community. Therefore, this qualifies as an AI Incident under the framework, as the AI system's use directly led to harm.[AI generated]

The article describes plausible future risks related to AI, such as financial market stress and economic downturns due to AI investment fluctuations, which could lead to harm but have not yet materialized. There is no direct or indirect evidence of an AI system causing harm or malfunctioning. Therefore, this event fits the definition of an AI Hazard, as it plausibly could lead to an AI Incident in the future but does not describe an actual incident or harm currently occurring.[AI generated]

An AI system (Google's Gemini) was explicitly used to generate answers from stolen exam papers, directly enabling cheating. This use of AI caused harm by violating the fairness and trust in the exam process, which is a harm to communities and a breach of obligations under applicable law protecting fair competition and rights. Therefore, this event qualifies as an AI Incident due to the realized harm caused by the AI system's misuse in an official exam setting.[AI generated]

The involvement of AI systems is explicit in the use of AI-generated images and AI-based advertising and search result summaries that facilitate recruitment. The harm is direct and materialized, as vulnerable individuals are being targeted and recruited for war, which is a violation of human rights and causes harm to communities. The use of AI-generated content and AI-driven ad targeting is central to the harm, making this an AI Incident rather than a hazard or complementary information. The event is not merely a report or update but documents ongoing exploitation and harm caused by AI systems.[AI generated]

The article involves an AI system indirectly through the data centre supporting AI and cloud services. The concerns raised are about potential environmental harms (water use, electricity consumption, climate impact) that could plausibly lead to harm in the future. No actual injury, rights violation, or environmental damage has been reported yet. The protests and concerns reflect a credible risk of harm from the AI infrastructure expansion, fitting the definition of an AI Hazard. It is not Complementary Information because the main focus is on the potential harm and protests, not on responses or updates to a past incident. It is not Beneficial Use or Unrelated as AI involvement and plausible harm are central.[AI generated]

The article explicitly mentions AI-generated content (AI short dramas) using faces sourced without proper authorization, constituting a violation of personal rights and intellectual property. The AI system's use in generating these virtual characters directly leads to harm through unauthorized exploitation of individuals' likenesses. The harm is materialized, as evidenced by legal rulings and content takedowns. Hence, this event meets the criteria for an AI Incident due to realized violations of rights caused by AI system use and misuse.[AI generated]

The event involves AI systems in the context of military decision-making and misinformation risks, which could plausibly lead to significant harm such as military conflict or misuse of AI for harmful purposes. Since no actual harm has been reported but credible risks and near incidents are described, this qualifies as an AI Hazard. The article does not report a realized AI Incident, nor is it merely complementary information or unrelated news. Therefore, the classification as AI Hazard is appropriate.[AI generated]

The article explicitly mentions AI systems used for automated license plate recognition and AI-assisted report writing, indicating AI system involvement. The event stems from the use and expansion of these AI systems by the LAPD. Although privacy and data sharing concerns are noted, no direct or indirect harm such as violations of rights or other harms have been reported as having occurred. The event involves a pilot program and contract renewal that could plausibly lead to harms related to privacy, surveillance overreach, or data misuse in the future. Hence, it fits the definition of an AI Hazard, as the AI systems' deployment could plausibly lead to incidents involving harm to rights or communities, but no harm has yet materialized.[AI generated]

The event involves AI systems embedded in smart glasses that record and analyze audio-visual data. The reported harm includes unauthorized recording of individuals without consent, including sensitive situations, and subsequent processing of this data by AI systems, which constitutes a violation of privacy rights and legal obligations. The presence of lawsuits and user complaints confirms that harm has materialized. The AI system's use is directly linked to these harms, fulfilling the criteria for an AI Incident under violations of human rights and legal protections. The article also discusses broader societal and industry responses, but the primary focus is on the realized harm caused by the AI system's use.[AI generated]

The AI system was used to identify a person in a video, but it incorrectly attributed the identity of Flavia Broffoni, who was not involved. This misidentification led to widespread dissemination of false information and violent threats against her, including death and sexual violence threats. The harm is direct and significant, involving violations of personal rights and threats to physical safety. The AI's role is pivotal as the misidentification by the AI system triggered the viral spread and subsequent harassment. Hence, this is an AI Incident rather than a hazard or complementary information.[AI generated]

The event involves AI systems (autonomous vehicles) and their planned use, but no actual harm or malfunction has occurred yet. The article focuses on the anticipated deployment, regulatory development, and technological considerations, indicating a plausible future scenario where AI systems could lead to incidents or hazards. However, since no harm or incident has materialized, and the article does not report any near misses or accidents, it fits the definition of an AI Hazard, reflecting a credible potential for future harm once these systems are operational.[AI generated]

The article explicitly describes an AI system (LynksOS) designed for autonomous control of defense vehicles and critical subsystems, involving AI for perception, planning, and decision support. The system is under development and funded by a government ministry, with plans for deployment in autonomous defense platforms. No actual harm or incident is reported; the focus is on development and capabilities. However, autonomous defense systems with AI have inherent risks of causing harm if malfunctioning or misused, which makes this a plausible future hazard. Thus, the event is best classified as an AI Hazard rather than an Incident or Complementary Information.[AI generated]

The event involves an AI system designed to manage content and user interaction dynamically during gameplay, which qualifies as AI involvement. However, the patent is only a proposal and has not been implemented or used, so no direct or indirect harm has occurred. The potential harm is to user experience and community satisfaction, but this remains speculative at this stage. Therefore, this event represents a plausible future risk of harm if implemented, qualifying it as an AI Hazard rather than an AI Incident or Complementary Information. It is not Beneficial Use since the AI is not countering external harm, nor is it unrelated.[AI generated]

The article does not report any actual harm caused by AI systems or data centres but warns of plausible future harms related to environmental resource depletion and community impacts due to the rapid build-up of AI-driven data centres. The AI involvement is clear as these data centres support AI services. The concerns about water stress, power consumption, noise, and land use represent credible risks that could lead to AI Incidents if unaddressed. Therefore, this is an AI Hazard, not an Incident or Complementary Information, as the harms are potential rather than realized.[AI generated]