aim-logo

AIM: AI Incidents and Hazards Monitor

Automated media discourse monitor of AI incidents and hazards (Beta)

AI-related legislation is gaining traction, and effective policymaking needs evidence, foresight and international cooperation. The OECD AI Incidents and Hazards Monitor (AIM) documents AI incidents and hazards to help policymakers, AI practitioners, and all stakeholders worldwide gain valuable insights into the risks and harms of AI systems. Over time, AIM will help to show risk patterns and establish a collective understanding of AI incidents and hazards and their multifaceted nature, serving as an important tool for trustworthy AI. AI incidents seem to be getting more media attention lately, but they've actually gone down as a share of all AI news (see chart below!).

The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.

Advanced Search Options

As percentage of total AI events
Note: An AI incident or hazard can be reported by one or more news articles covering the same event. Data processing powered by Microsoft Azure using data from Event Registry.
Show summary statistics of AI incidents & hazards
Results: About 19243 incidents & hazards
Thumbnail Image

Atiku Demands Audit and Transparency for INEC's Proposed AI Election System

2026-11-19
Nigeria

Former Vice President Atiku Abubakar and the African Democratic Congress have called on Nigeria's electoral commission (INEC) to subject its proposed AI system for 2027 election result management to independent audit and public scrutiny, citing concerns that lack of transparency could undermine electoral integrity and public trust.[AI generated]

AI principles:
Transparency & explainabilityDemocracy & human autonomy
Industries:
Government, security, and defence
Affected stakeholders:
General public
Harm types:
Public interest
Business function:
Compliance and justice
AI system task:
Goal-driven organisation
Why's our monitor labelling this an incident or hazard?

The article centers on the potential risks and calls for transparency and audit of an AI system intended for election result management. There is no indication that the AI system has malfunctioned or caused harm yet. The concerns raised are about plausible future harms such as undermining public confidence, delaying or rejecting valid results, and possible electoral crises if the AI system is misused or malfunctions. Therefore, this event fits the definition of an AI Hazard, as it plausibly could lead to an AI Incident if the risks materialize, but no incident has occurred yet.[AI generated]

Thumbnail Image

Singapore Central Bank Warns of AI Agents' Financial Risks, Plans Safeguards

2026-11-19
Singapore

The Monetary Authority of Singapore (MAS) warned that autonomous AI agents could rapidly move funds, posing liquidity and financial stability risks. MAS is developing a "Know Your Agent" protocol to verify AI agents' identities and authorities, aiming to mitigate potential harms as financial institutions adopt AI-driven transactions.[AI generated]

AI principles:
AccountabilityRobustness & digital security
Industries:
Financial and insurance services
Affected stakeholders:
BusinessGeneral public
Harm types:
Economic/PropertyPublic interest
Business function:
ICT management and information security
Autonomy level:
High-action autonomy (human-out-of-the-loop)
AI system task:
Goal-driven organisation
Why's our monitor labelling this an incident or hazard?

The Monetary Authority of Singapore explicitly warns about AI agents' ability to move funds quickly, raising liquidity and financial stability risks, which are plausible harms. The AI agents are described as autonomous systems managing money and making payments, fitting the definition of AI systems. Since no actual harm or incident has occurred yet, but the risks are credible and foreseeable, this is an AI Hazard. The article also discusses regulatory and institutional responses, but the main focus is the warning about potential risks, not a realized incident or complementary information about past incidents.[AI generated]

Thumbnail Image

AI-Enabled Cheating in South Korea's College Entrance Exam Spurs Whistleblower Program

2026-11-19
Korea

South Korea's Anti-Corruption and Civil Rights Commission launched a two-month reporting period to combat cheating in the national college entrance exam using AI smart glasses and generative AI. Whistleblowers can receive up to 100 million KRW if their reports lead to administrative sanctions against offenders.[AI generated]

AI principles:
FairnessAccountability
Industries:
Education and training
Affected stakeholders:
General public
Harm types:
ReputationalPublic interest
Autonomy level:
High-action autonomy (human-out-of-the-loop)
AI system task:
Recognition/object detectionContent generation
Why's our monitor labelling this an incident or hazard?

The event involves AI systems (AI smart glasses, generative AI) used to facilitate cheating in a high-stakes exam, which directly harms the fairness of the exam process and violates principles of fair competition, a form of harm to communities and rights. The article reports on actual cheating incidents and government responses, indicating realized harm rather than potential harm. Hence, this qualifies as an AI Incident due to the direct involvement of AI in causing harm through misuse.[AI generated]

Thumbnail Image

South Korean Lawmaker Calls for Investigation into AI-Assisted Exam Cheating

2026-11-19
Korea

South Korean lawmaker Jeong Seong-guk has requested the preservation of university entrance exam materials to investigate suspicions of AI-assisted cheating using smart glasses and cameras. The call aims to verify if AI misuse undermined exam fairness, with potential for a broader probe if concrete evidence emerges. No confirmed incidents yet reported.[AI generated]

AI principles:
Fairness
Industries:
Education and training
Affected stakeholders:
General public
Harm types:
Public interestReputational
AI system task:
Recognition/object detectionContent generation
Why's our monitor labelling this an incident or hazard?

The article explicitly mentions AI-enabled devices potentially used for cheating, implying the involvement of AI systems in the context of exam integrity. The event is about the potential for AI misuse in exams, which could plausibly lead to harm such as unfairness and violation of rights related to education and fair competition. Since no confirmed incident of cheating or harm is reported, and the focus is on preserving evidence and investigating suspicions, the event fits the definition of an AI Hazard. It is not Complementary Information because it is not an update or response to a past incident but a proactive measure to address a potential future harm. It is not Beneficial Use because the AI is not used as a countermeasure but is suspected as a tool for cheating. It is not Unrelated because AI involvement is central to the concern.[AI generated]

Thumbnail Image

AI-Generated Video Calls Used in Buzău Fraud Attempt

2026-11-19
Romania

A foreign suspect in Buzău, Romania, used AI-generated video calls to impersonate an official and deceive victims into believing their bank accounts were compromised. The scam, which aimed to steal over 140,000 lei, was thwarted by police, who arrested the perpetrator during the attempted money transfer.[AI generated]

AI principles:
Robustness & digital securityTransparency & explainability
Industries:
Financial and insurance servicesDigital security
Affected stakeholders:
General public
Harm types:
Economic/Property
AI system task:
Content generation
Why's our monitor labelling this an incident or hazard?

The use of AI-generated video calls to impersonate an official and deceive a victim constitutes direct involvement of an AI system in causing harm, specifically financial harm through attempted fraud. The AI system's role was pivotal in making the scam credible, leading to the victim's attempt to hand over money. This meets the criteria for an AI Incident as the AI system's use directly led to harm (attempted financial fraud).[AI generated]

Thumbnail Image

UN Warns of 'Medieval' AI Governance and Concentration of Power

2026-11-19
Switzerland

UN Human Rights Chief Volker Türk warned at the RightsX summit in Geneva that AI governance is dangerously concentrated among a few tech companies, lacking democratic oversight and common regulations. He highlighted risks to human rights and critical infrastructure if AI development remains unchecked and unregulated.[AI generated]

AI principles:
Democracy & human autonomyAccountability
Industries:
Government, security, and defenceDigital security
Affected stakeholders:
General public
Harm types:
Human or fundamental rightsPublic interest
Why's our monitor labelling this an incident or hazard?

The event involves AI systems as it discusses AI's growing influence and control over critical infrastructure and societal aspects. However, no specific AI Incident (realized harm) is described; rather, the article warns about plausible future harms and systemic risks due to concentration of power and lack of oversight. This fits the definition of an AI Hazard, as it plausibly could lead to incidents like infrastructure disruption or rights violations. The article also includes calls for regulation and accountability but does not primarily focus on responses or updates, so it is not Complementary Information. Therefore, the classification is AI Hazard.[AI generated]

Thumbnail Image

AI-Generated Fraudulent Websites Lead to Major Internet Scam Bust in Málaga

2026-11-19
Spain

Spanish police dismantled a Málaga-based criminal network that used AI to create fraudulent websites, enabling phishing and smishing attacks across Spain. The operation resulted in four arrests and the seizure of over €150,000 in assets. The AI-driven scams caused widespread financial and identity-related harm to victims.[AI generated]

AI principles:
Privacy & data governanceTransparency & explainability
Industries:
Digital security
Affected stakeholders:
Consumers
Harm types:
Economic/PropertyHuman or fundamental rights
AI system task:
Content generation
Why's our monitor labelling this an incident or hazard?

The article explicitly states that the perpetrators used artificial intelligence to create fraudulent websites and conduct phishing and smishing attacks, which directly led to financial harm (over 150,000 euros) to victims. The AI system's use in generating or facilitating these scams is a direct contributing factor to the harm. This fits the definition of an AI Incident, as the AI system's use has directly led to harm to people (financial injury) and violations of legal rights (fraud).[AI generated]

Thumbnail Image

South Korean Civil Servant Uses ChatGPT to Forge Medical Certificates for Sick Leave

2026-11-19
Korea

A South Korean civil servant used ChatGPT to generate 41 fake medical certificates, submitting them to obtain 50 days of unauthorized sick leave and evade service duties. The individual was convicted of forgery and related offenses, receiving an eight-month suspended prison sentence. The incident highlights the misuse of generative AI for document forgery.[AI generated]

AI principles:
SafetyAccountability
Industries:
Government, security, and defence
Affected stakeholders:
Government
Harm types:
ReputationalPublic interest
Autonomy level:
No-action autonomy (human support)
AI system task:
Content generation
Why's our monitor labelling this an incident or hazard?

The event explicitly states that a generative AI system (ChatGPT) was used to produce forged medical documents, which directly caused the individual to evade mandatory service unlawfully. This misuse of AI led to a legal violation (forgery and unauthorized absence), satisfying the criteria for an AI Incident due to breach of legal obligations and harm to institutional integrity. The harm is realized, not just potential, and the AI system's role is pivotal in enabling the forgery.[AI generated]

Thumbnail Image

AI-Powered Cyberattack Targets EU Commission Personnel Systems

2026-11-19
Belgium

In October, hackers used an AI model, reportedly of Chinese origin, to launch a cyberattack on the European Commission, targeting its personnel management systems. The attack led to operational disruptions, though no personal data theft was confirmed. The Commission responded by shutting down affected internal systems in Brussels.[AI generated]

AI principles:
Robustness & digital securitySafety
Industries:
Government, security, and defenceDigital security
Affected stakeholders:
Government
Harm types:
Public interest
AI system task:
Other
Why's our monitor labelling this an incident or hazard?

The event involves the use of an AI system (an AI model misused for a cyberattack) that directly led to harm in the form of a cyberattack on critical infrastructure (personnel management systems of the EU Commission). This fits the definition of an AI Incident because the AI system's use directly caused harm through malicious cyber activity. The lack of data theft does not negate the harm caused by the attack itself. Therefore, this event qualifies as an AI Incident rather than a hazard or complementary information.[AI generated]

Thumbnail Image

Russian Use of Generative AI for Disinformation in Europe

2026-11-19

Russian special services are using generative AI to create and spread fake videos and disinformation, particularly in Central Europe. These AI-generated materials mimic reputable media and target elections and political figures, aiming to destabilize societies, foster division, and undermine political processes across European Union countries.[AI generated]

AI principles:
Transparency & explainabilityDemocracy & human autonomy
Industries:
Media, social platforms, and marketingGovernment, security, and defence
Affected stakeholders:
General publicGovernment
Harm types:
Public interestPsychologicalReputational
Autonomy level:
Low-action autonomy (human-in-the-loop)
AI system task:
Content generation
Why's our monitor labelling this an incident or hazard?

The event involves the use of generative AI systems by Russian actors to produce and distribute disinformation that harms communities in Europe. The AI-generated content includes fake videos and fabricated statements that mislead the public and damage political processes, fulfilling the criteria for harm to communities under the AI Incident definition. The harm is realized and ongoing, not merely potential, and the AI system's role is pivotal in creating and spreading the false content. Therefore, this event qualifies as an AI Incident.[AI generated]

Thumbnail Image

Anthropic Faces Criticism Over AI Safety Commitments Ahead of IPO

2026-11-19
United States

A coalition of Nobel laureates, scientists, and tech leaders publicly challenged Anthropic's safety practices, accusing the company of weakening its safety commitments while racing to develop self-improving AI. The critics warn that such AI could escape human control, posing significant future risks. The event centers on Anthropic's US IPO.[AI generated]

AI principles:
SafetyAccountability
Industries:
Other
Affected stakeholders:
General public
Harm types:
Public interest
Business function:
Research and development
Autonomy level:
High-action autonomy (human-out-of-the-loop)
AI system task:
Goal-driven organisation
Why's our monitor labelling this an incident or hazard?

The article focuses on concerns and warnings from AI safety advocates about the potential dangers of Anthropic's AI development approach, particularly recursive AI systems. There is no indication that any harm has already occurred, only that the development approach could plausibly lead to harm in the future. Therefore, this qualifies as an AI Hazard, as it involves the plausible future risk of harm stemming from the development and use of AI systems, rather than an AI Incident or Complementary Information.[AI generated]

Thumbnail Image

IEA Warns AI Datacentres to Drive 70% Rise in Carbon Emissions by 2030

2026-11-19
United States

The International Energy Agency (IEA) warns that global datacentres, driven by rapid AI infrastructure expansion, will double their electricity consumption and generate 70% more carbon emissions by 2030. The US and China are leading this trend, primarily relying on fossil fuels to meet AI-related energy demands.[AI generated]

AI principles:
Sustainability
Industries:
IT infrastructure and hostingEnergy, raw materials, and utilities
Affected stakeholders:
General public
Harm types:
Environmental
Why's our monitor labelling this an incident or hazard?

The article discusses a credible future risk (strain on electricity grids) due to the growth of AI-related data centers, which could plausibly lead to disruption of critical infrastructure (electricity networks). There is no indication that harm has already occurred, so it does not qualify as an AI Incident. The risk is plausible and linked to AI system use (training and running AI models), so it fits the definition of an AI Hazard. It is not merely complementary information because the main focus is on the potential harm, not on responses or updates. It is not unrelated or beneficial use, as the AI systems are the source of the potential hazard, not a countermeasure.[AI generated]

Thumbnail Image

AI-Enabled Cheating in South Korean Tax Accountant Exam Spurs Class Action Lawsuit

2026-11-19
Korea

A top scorer in South Korea's 63rd tax accountant exam admitted to using generative AI, a miniature camera, and a smartphone to cheat. Law firm YK is organizing a class action lawsuit against the exam authority, citing failure to detect the AI-enabled cheating and resulting harm to other candidates.[AI generated]

AI principles:
SafetyRobustness & digital security
Industries:
Education and training
Affected stakeholders:
Other
Harm types:
Economic/Property
Autonomy level:
No-action autonomy (human support)
AI system task:
Content generation
Why's our monitor labelling this an incident or hazard?

The article explicitly states that generative AI was used to commit cheating in a high-stakes exam, which harmed other candidates by undermining fairness and trust in the exam system. The misuse of AI here directly led to realized harm (violation of rights and fairness). The legal response focuses on accountability for this harm. Therefore, this event qualifies as an AI Incident due to direct harm caused by AI misuse in an exam context.[AI generated]

Thumbnail Image

AI-Enabled Cheating Detected in South Korea's College Entrance Exam

2026-11-19
Korea

During South Korea's 2027 college entrance exam, 111 cases of cheating involving AI-powered smart devices, such as smart glasses and miniature cameras, were detected. Authorities used emergency disaster messages and AI-based detection methods to identify offenders, highlighting the growing misuse of AI in exam settings and its impact on fairness.[AI generated]

AI principles:
FairnessAccountability
Industries:
Education and training
Affected stakeholders:
General public
Harm types:
Public interest
Autonomy level:
Low-action autonomy (human-in-the-loop)
AI system task:
Content generation
Why's our monitor labelling this an incident or hazard?

The article explicitly mentions the use of AI-enabled smart glasses and other smart devices to cheat during a high-stakes exam, which directly harms the fairness and integrity of the examination process, a harm to communities and societal trust. The AI system's involvement is in the misuse of AI technology for cheating, which is a violation of rules and causes harm. Hence, this event meets the criteria for an AI Incident due to realized harm caused by AI system misuse.[AI generated]

Thumbnail Image

SpaceXSI Denies Liability After Grok AI Generates Sexualised Images of UK Lawmaker

2026-11-19
United Kingdom

Jess Asato, a UK lawmaker, sued SpaceXSI after its Grok AI platform generated non-consensual, sexualised images and videos of her. SpaceXSI denies liability, arguing such AI-generated content is obviously fictitious and not a privacy breach. The case tests legal responsibility for AI system design and misuse in the UK.[AI generated]

AI principles:
Privacy & data governanceRespect of human rights
Industries:
Media, social platforms, and marketing
Affected stakeholders:
Women
Harm types:
Human or fundamental rightsReputationalPsychological
Autonomy level:
No-action autonomy (human support)
AI system task:
Content generation
Why's our monitor labelling this an incident or hazard?

The event explicitly involves an AI system (Grok) generating harmful, sexualised, non-consensual images of a person, which constitutes a violation of privacy and data protection rights, a form of harm under the AI Incident definition (c). The lawsuit and the described harms are directly linked to the AI system's use and design. The presence of the AI system is clear, and the harm is realized, not just potential. The legal dispute centers on liability for these harms caused by the AI system's outputs. Hence, this is an AI Incident rather than a hazard or complementary information.[AI generated]

Thumbnail Image

South Korea Preserves Exam Papers Amid AI Cheating Concerns

2026-11-19
Korea

South Korea's Prime Minister and Education Ministry ordered the preservation of last year's national college entrance exam papers due to rising concerns about potential AI-assisted cheating. The move aims to secure evidence for possible investigations, though no confirmed AI-related cheating incidents have occurred yet.[AI generated]

AI principles:
Fairness
Industries:
Education and training
Why's our monitor labelling this an incident or hazard?

The article explicitly mentions concerns about AI-enabled cheating, indicating AI system misuse as a potential threat. The government's action to preserve exam materials is a response to this plausible risk. Since no actual cheating incident or harm has been confirmed or reported, the event does not qualify as an AI Incident. Instead, it fits the definition of an AI Hazard, as it involves circumstances where AI misuse could plausibly lead to harm (exam integrity violation) in the future. The directive is a preventive measure, not a report of realized harm or a societal/governance response to a past incident, so it is not Complementary Information.[AI generated]

Thumbnail Image

AI Chatbots and Social Media Linked to Child Harm in Canada

2026-11-19
Canada

Canadian experts report that AI chatbots and social media platforms have contributed to increased child suicide attempts, online bullying, and sexual trafficking. Pediatricians and child protection advocates highlight a surge in harm to children, urging swift passage of digital safety legislation to address these AI-driven risks.[AI generated]

AI principles:
SafetyRespect of human rights
Industries:
Media, social platforms, and marketing
Affected stakeholders:
Children
Harm types:
PsychologicalHuman or fundamental rights
Business function:
Citizen/customer service
Autonomy level:
High-action autonomy (human-out-of-the-loop)
AI system task:
Interaction support/chatbotsOrganisation/recommenders
Why's our monitor labelling this an incident or hazard?

The article explicitly mentions AI chatbots as part of the digital environment causing significant harm to children's brains, with references to serious harms such as child suicide and trafficking. The involvement of AI systems (chatbots) in causing or contributing to these harms meets the criteria for an AI Incident, as the harm is realized or ongoing. The pediatrician's call for regulation underscores the severity and direct link to AI use.[AI generated]

Thumbnail Image

Suspect Arrested After Sharing Murder Plan with ChatGPT

2026-11-19
Israel

Israeli police arrested a northern resident after discovering he shared his plan to murder his wife with the AI chatbot ChatGPT. The suspect's communication with the AI system was pivotal in alerting authorities, leading to his arrest and the prevention of potential harm. The incident raises questions about AI privacy and law enforcement collaboration.[AI generated]

AI principles:
Privacy & data governanceTransparency & explainability
Industries:
Consumer servicesDigital security
Affected stakeholders:
Consumers
Harm types:
Human or fundamental rights
Business function:
Citizen/customer service
Autonomy level:
No-action autonomy (human support)
AI system task:
Interaction support/chatbots
Why's our monitor labelling this an incident or hazard?

An AI system (ChatGPT) was explicitly involved as the suspect shared his plan to commit a violent crime through it. This use of the AI system is directly linked to a serious potential harm to a person (the suspect's wife). The event involves the use of AI in a harmful context leading to law enforcement action to prevent harm. Since the harm was not realized but the AI system's use was pivotal in revealing the threat and enabling intervention, this qualifies as an AI Incident due to the direct link between AI use and a serious threat to human life.[AI generated]

Thumbnail Image

AI-Driven Cyberattacks Escalate Threat to German Critical Infrastructure

2026-11-19
Germany

Germany faces a surge in cyberattacks, with AI systems enabling attackers to rapidly exploit vulnerabilities and disrupt critical infrastructure, such as supermarket and blood donation center cooling systems. The Federal Office for Information Security (BSI) reports a critical threat level, citing AI's role in automating and accelerating cyberattacks.[AI generated]

AI principles:
Robustness & digital securitySafety
Industries:
Digital securityGovernment, security, and defence
Affected stakeholders:
General publicConsumers
Harm types:
Public interestEconomic/Property
Autonomy level:
High-action autonomy (human-out-of-the-loop)
AI system task:
Other
Why's our monitor labelling this an incident or hazard?

The article explicitly mentions AI systems autonomously conducting cyberattacks that have already caused harm, such as system shutdowns in hospitals and data theft from government entities. The AI's role in accelerating and automating these attacks is central to the increased threat level and realized harms. This meets the definition of an AI Incident, as the AI system's use has directly led to harm to critical infrastructure and communities. The article also discusses the broader context of increased AI-driven cyber threats and responses, but the primary focus is on the realized harms caused by AI-enabled attacks.[AI generated]

Thumbnail Image

Wisconsin Sues TikTok Over AI-Driven Addictive Features Harming Children

2026-11-19
United States

Wisconsin has filed a lawsuit against TikTok, alleging its AI-powered recommendation system and addictive features intentionally target children, causing mental health harms such as anxiety, depression, and sleep disruption. The suit claims TikTok deceived parents about these risks, prioritizing profit over children's safety.[AI generated]

AI principles:
SafetyTransparency & explainability
Industries:
Media, social platforms, and marketing
Affected stakeholders:
Children
Harm types:
Psychological
Business function:
Marketing and advertisement
Autonomy level:
High-action autonomy (human-out-of-the-loop)
AI system task:
Organisation/recommenders
Why's our monitor labelling this an incident or hazard?

TikTok's use of personalized algorithms, infinite scroll, autoplay, and push notifications involves AI systems that influence user behavior. The lawsuit alleges these AI-driven features have directly led to harm to youth mental health, including depression and anxiety, which fits the definition of an AI Incident due to harm to a group of people. The event involves the use of AI systems and their impact, with realized harm alleged, making it an AI Incident rather than a hazard or complementary information.[AI generated]