AIM: AI Incidents and Hazards Monitor

The incident involves AI systems (deepfake generators) directly causing financial harm (losses to individuals) through fraudulent investment schemes on social media. The harm has occurred and is ongoing, making this an AI Incident rather than a potential hazard or purely a governance update.[AI generated]

The event describes actual deployment of AI-based facial recognition leading to non-consensual identification and doxxing of individuals, constituting a violation of fundamental privacy rights. This is a realized harm directly linked to misuse of an AI system.[AI generated]

The robot dog Spot is an AI system used by the police in a real-world scenario where it identified and retrieved a weapon from a suspect, directly contributing to a peaceful resolution and preventing harm. This constitutes direct involvement of an AI system in an event with significant implications for public safety and human rights (concerns about surveillance and militarization). The article also references prior controversies and public reactions, but the core event is the AI system's operational use leading to harm prevention, which fits the definition of an AI Incident. The concerns about future misuse or militarization are noted but secondary to the realized impact of the AI system in this case.[AI generated]

This report details actual, realized fraudulent activities—scammers are using AI to create convincing impersonations and solicit money or personal data, leading to financial losses (harm to property). Thus it qualifies as an AI Incident.[AI generated]

Pegasus spyware is an AI-enabled system used for surveillance and data extraction. The unauthorized use of this system to extract a large volume of private messages and manipulate them for political discreditation directly leads to violations of privacy and human rights. The event involves the use and misuse of an AI system causing harm, fitting the definition of an AI Incident under violations of human rights and harm to communities. Therefore, this event is classified as an AI Incident.[AI generated]

TikTok uses AI algorithms for content recommendation and personalized advertising, which are central to the platform's operation. The lawsuit alleges that TikTok shared children's personal data without consent and failed to protect privacy, leading to violations of legal rights and exposing minors to risks. This constitutes harm to rights under applicable law, fulfilling the criteria for an AI Incident. The AI system's use in data processing and ad targeting directly contributed to the harm and legal action. Hence, the event is classified as an AI Incident.[AI generated]

Nokia’s intelligence report shows that cybercriminals are actively using generative AI and automation to carry out distributed denial-of-service attacks and other intrusions against telecom operators, causing actual harm and disruption to critical infrastructure. This is a realized event where an AI system’s use has directly led to significant harm to networks, fitting the definition of an AI Incident.[AI generated]

The event involves an AI system (an AI transcription service) that malfunctioned or was misused by automatically sharing private meeting transcripts containing confidential information. This directly caused harm by exposing sensitive business details and influencing a business decision negatively. This fits the definition of an AI Incident because the AI system's use directly led to a violation of confidentiality and harm to the involved parties' interests.[AI generated]

The article describes an algorithmic video‐surveillance system already trialed under the JOP law and now slated for permanent roll-out. It details an AI system’s intended use and the plausible human rights and privacy harms that could follow, but does not report any realized incident. This fits the definition of an AI Hazard.[AI generated]

The article describes the initiation of an AI system development project with potential defense applications but does not report any actual harm, malfunction, or misuse of the AI system. The event is about the start of a project and funding awarded, which could plausibly lead to future impacts but does not itself constitute an incident or hazard. It is not a complementary information piece about an existing incident or hazard, nor is it unrelated to AI. Therefore, it fits best as an AI Hazard because the development of such AI systems for tactical spectrum classification in defense could plausibly lead to future harms or risks, but no harm has yet occurred or been reported.[AI generated]

The event involves an AI system explicitly: the automated driving system (ADAS) in the vehicle. The driver misused the system by tricking it to allow hands-free driving while being inattentive and asleep, which directly created a hazardous situation on a highway. The article also references multiple past accidents caused by similar AI systems, some fatal, establishing a pattern of harm linked to the use and misuse of these AI systems. The misuse and malfunction of the AI system have directly or indirectly led to harm or risk of harm to human life, fulfilling the criteria for an AI Incident. Although the specific incident described did not result in an accident, the direct risk to life and the history of actual accidents caused by such systems justify classification as an AI Incident rather than a mere hazard.[AI generated]

The event involves an AI system concept (fully autonomous aircraft) that could plausibly lead to harm in the future, such as injury to passengers or disruption of aviation infrastructure, but the plane is still at the conceptual stage with no deployment or malfunction reported. Therefore, it qualifies as an AI Hazard due to the credible risk posed by the intended use of AI in autonomous passenger flight, but not an AI Incident since no harm has occurred.[AI generated]

Although the robot dog has not yet caused harm, it embodies an AI-powered lethal capability whose development and testing plausibly lead to significant future harms (e.g., unintended civilian casualties, escalation of force, rights violations). Therefore, it constitutes an AI Hazard rather than an incident or mere product announcement.[AI generated]

The incident involves the malicious use of an AI voice‐cloning system to commit fraud. The father was deceived into believing his son was in legal trouble and wired money for bail. This constitutes direct harm (financial loss) resulting from the AI system’s outputs, fitting the definition of an AI Incident.[AI generated]

The article describes how a deployed AI system directly caused harm—wrongful detention—by misidentifying an individual in criminal investigations. This constitutes a realized harm to the person’s rights and liberty, so it is an AI Incident.[AI generated]

The article focuses on the potential risks and security concerns arising from the use of generative AI by employees without formal training, which could plausibly lead to harms such as data breaches or legal violations. However, no actual harm or incident is reported. Therefore, this qualifies as an AI Hazard because it describes circumstances where AI use could plausibly lead to harm, but no direct or indirect harm has yet materialized. It is not Complementary Information because it is not updating or responding to a previously reported incident, nor is it unrelated as it clearly involves AI systems and their use in the workplace.[AI generated]

The article describes a prospective policy—requiring banks to automatically flag “suspicious behaviour” via algorithmic/AI systems for benefit fraud—but no actual system has yet been deployed or caused harm. The focus is on the potential for large-scale, suspicionless financial surveillance and punitive errors, making it a credible future risk rather than a realized incident.[AI generated]

The article focuses on the development and planned use of AI and autonomous systems in military operations, highlighting strategic intentions and capabilities rather than any specific AI-related harm or malfunction. While the integration of AI in military systems inherently carries risks, the article does not describe any actual AI Incident or harm caused by AI. The potential for future harm exists given the military applications and geopolitical tensions, but this remains a plausible risk rather than a realized event. Therefore, the event is best classified as an AI Hazard, reflecting the credible potential for AI-related harm in the future due to the deployment and use of AI-enabled military systems.[AI generated]

Tesla’s FSD is an AI-driven driver-assist system whose malfunction (failure to register an SUV in a blind spot) directly contributed to a collision with minor injuries and a totaled car. The incident demonstrates realized harm from an AI system in use, making this an AI Incident.[AI generated]