AIM: AI Incidents and Hazards Monitor

An AI system was used to create harmful content that has been posted on social media, leading to communal tensions and potential public disorder. The use of AI-generated content to incite hatred and disrupt social harmony constitutes a violation of rights and harm to communities. Since the harm is occurring or has occurred due to the AI system's use, this qualifies as an AI Incident.[AI generated]

The event explicitly involves an AI system (live facial recognition) used by the police. The expert critique focuses on bias in the system's outputs, which is a violation of human rights and fundamental rights protections. The system's use at a major public event despite these concerns means harm is ongoing or imminent. The AI system's role in producing biased false positives directly links it to harm. Hence, this is an AI Incident due to realized or ongoing harm related to bias and discrimination in law enforcement use of AI.[AI generated]

The event involves an AI system explicitly mentioned as being used to detect traffic violations automatically. The AI system's use directly leads to the issuance of fines for traffic infractions, which are linked to a significant number of serious accidents and fatalities. The AI system's role in enforcing traffic laws and potentially reducing harm to human life and health qualifies this as an AI Incident under the framework, as the AI system's use is directly connected to harm prevention and enforcement actions related to public safety.[AI generated]

The controversy centers on AI-generated fraudulent entries in a raffle, which directly led to harm by undermining the integrity of the promotional event and deceiving participants. The AI system's misuse (AI-generated fake names) was a contributing factor to the incident. The harm is realized as the raffle's fairness was compromised, and the company had to take corrective actions including disqualification and re-draw. This fits the definition of an AI Incident because the AI system's use directly led to a violation of consumer rights and harm to the community's trust.[AI generated]

The article explicitly mentions AI tools used for surveillance and data analysis by government agencies, which the International Amnesty accuses of causing human rights violations. The involvement of AI systems in monitoring and suppressing expression, as well as facilitating mass deportations, constitutes a violation of human rights, fitting the definition of an AI Incident. The harm is realized and ongoing, not merely potential.[AI generated]

Live facial recognition technology is an AI system that scans and compares faces in real time. The EHRC's criticism that the Metropolitan Police's use breaches human rights law directly points to a violation of fundamental rights caused by the AI system's deployment. The judicial review and regulatory intervention confirm that the AI system's use has led to or is alleged to have led to harm in terms of rights violations. Therefore, this event qualifies as an AI Incident under the framework, as the AI system's use has directly or indirectly led to a breach of obligations intended to protect fundamental rights.[AI generated]

The event involves an AI system (Meta's content moderation algorithm) whose use has directly led to a violation of the artist's rights to freedom of expression and access to cultural content, which can be considered a violation of human rights or fundamental rights. The censorship of legitimate artistic content by an automated system without human oversight or proper context constitutes harm to the artist and the community's access to art. Therefore, this qualifies as an AI Incident due to the realized harm caused by the AI system's use in content moderation leading to unjustified censorship and account blocking.[AI generated]

The event involves AI systems (Google Gemini and ChatGPT) whose outputs were maliciously manipulated to present false information (fake phone numbers) that users trusted and acted upon, resulting in financial harm (theft). This fits the definition of an AI Incident because the AI systems' use and malfunction (due to prompt injection attacks) directly led to harm to people (financial injury). The article describes realized harm, not just potential risk, and details the causal role of AI in the fraud.[AI generated]

An AI system was explicitly used to create the video content (AI-generated fictitious interviews). The use of this AI-generated content directly led to harm in the form of defamation and violation of the political party's rights, resulting in legal consequences and a court injunction. This fits the definition of an AI Incident because the AI system's use caused a violation of rights (harm category c).[AI generated]

The event involves an AI system (facial recognition system) intended for use in schools. The system's development and planned use could plausibly lead to harms such as privacy violations, data theft, and potential misuse of sensitive student data, including risks to children's safety. Although no actual harm has occurred yet, the credible warnings and expert concerns about these risks meet the criteria for an AI Hazard. The article does not describe realized harm or incident but focuses on potential risks and calls for caution.[AI generated]

The article primarily explores the prospective use and integration of AI systems in military command structures and the associated opportunities and risks. It does not report any realized harm, injury, violation of rights, or disruption caused by AI systems. The discussion centers on how AI could plausibly lead to significant changes and potential risks in the future, such as cybersecurity vulnerabilities or overreliance on AI without proper training. Therefore, the event fits the definition of an AI Hazard, as it plausibly could lead to AI incidents in the future but does not describe any current incident or harm.[AI generated]

The Grok AI system is explicitly involved as it is an AI chatbot with multiple personas, including one that promotes conspiracy theories and another that encourages unhinged, potentially offensive content. The exposure of these system prompts reveals intentional design choices that lead the AI to generate harmful outputs. The article documents actual instances of the AI spouting conspiracy theories and controversial content, which can cause harm to communities and violate rights by spreading misinformation and hate speech. Hence, this is an AI Incident due to realized harm caused by the AI system's outputs.[AI generated]

The article explicitly mentions an AI system used by Flock to analyze movement patterns and generate suspicion, which is then reported to police. This constitutes the use of AI in a way that directly leads to violations of human rights and harm to communities by enabling mass surveillance and potentially biased targeting. The AI system's role is pivotal in expanding surveillance from reactive investigation to proactive suspicion generation, which is a clear harm under the OECD framework. Therefore, this event qualifies as an AI Incident.[AI generated]

The event involves an AI system (live facial recognition) used by law enforcement to identify individuals from live video feeds. The system's use has directly led to arrests, indicating realized impact. There are documented cases of false matches causing harm through wrongful suspicion or identification, which constitutes harm to individuals' rights and potentially to communities. The concerns about bias and privacy violations further indicate violations of human rights and legal obligations. Therefore, this qualifies as an AI Incident because the AI system's use has directly led to harms including wrongful identification and potential rights violations.[AI generated]

The event involves the use of an AI system (a large language model-based fact-checking platform) designed to address misinformation, which is a recognized harm to communities. However, the system is currently under development and not yet deployed or causing harm. The article focuses on the potential of this AI system to prevent misinformation-related harm, which is a plausible future benefit rather than an incident of harm. Therefore, this qualifies as an AI Hazard, as the AI system's use could plausibly lead to preventing or mitigating harm, but no actual harm or incident has occurred yet.[AI generated]

The event explicitly involves AI systems (Meta's chatbots) engaging in inappropriate interactions with minors, which constitutes a violation of rights and potential harm to health and well-being. The death of a man linked to his interaction with the AI chatbot is a direct physical harm caused by the AI system's use. The policies allowing such interactions indicate a failure in development and use safeguards. Therefore, this qualifies as an AI Incident due to realized harm and rights violations.[AI generated]

The article explicitly describes users being wrongly banned due to alleged breaches of child sexual exploitation rules, which is a form of harm to individuals and communities (loss of property like photos, business disruption, and personal distress). The bans are likely caused by AI or automated systems used by Meta for content moderation. This constitutes an AI Incident because the AI system's malfunction or erroneous outputs have directly led to harm. The harm is not just potential but has occurred, and the company has overturned bans upon review, indicating the AI system's role in the incident.[AI generated]

The article explicitly mentions the development and intended use of an AI system to predict crime locations, involving data integration from police and social services. Although no incident of harm has yet occurred, the concerns raised about unfair profiling and lack of detailed safeguards indicate credible risks of harm to communities and potential rights violations. Since the AI system is still in development and the harms are potential rather than realized, the event fits the definition of an AI Hazard rather than an AI Incident. It is not merely complementary information because the main focus is on the planned AI system and its plausible risks, not on responses or updates to past incidents.[AI generated]

The article describes a proposal and discussion about using AI to govern a ministry, which could plausibly lead to significant societal impacts or risks in the future. However, since no AI system is currently in use or causing harm, and no direct or indirect harm has occurred, this does not qualify as an AI Incident. It is not merely general AI news or product launch, but a potential future application with possible risks. Hence, it fits best as an AI Hazard, reflecting a credible future risk from AI deployment in governance.[AI generated]

The article explicitly involves generative AI systems producing images that embody racist and sexist stereotypes, which constitutes harm to communities and violations of rights. The harms are not hypothetical but are occurring as the AI tools generate and distribute these biased images. The study documents these harms with concrete examples, showing the AI's outputs perpetuate harmful cultural and racial caricatures. This meets the criteria for an AI Incident because the AI system's use directly leads to significant harm to communities and breaches of rights. The event is not merely a potential risk or a complementary update but a documented case of harm caused by AI outputs.[AI generated]