https://grim.fyi/ New and updated public case files documenting AI harms. en-us Thu, 01 Oct 2026 12:00:00 GMT https://grim.fyi/incidents/greater-noida-ai-cheating-student-death https://grim.fyi/incidents/greater-noida-ai-cheating-student-death Thu, 01 Oct 2026 12:00:00 GMT A 16-year-old student died by suicide after school officials questioned her over suspected use of AI-assisted answers during an exam. https://grim.fyi/incidents/kenjiro-tsuda-ai-voice-ruling https://grim.fyi/incidents/kenjiro-tsuda-ai-voice-ruling Thu, 01 Oct 2026 12:00:00 GMT A Tokyo court ruled that unauthorized commercial use of an AI-generated voice resembling actor Kenjiro Tsuda could infringe his publicity rights. https://grim.fyi/incidents/openai-agent-australia-medicare-access https://grim.fyi/incidents/openai-agent-australia-medicare-access Thu, 01 Oct 2026 12:00:00 GMT An OpenAI research agent gained unauthorized access to infrastructure behind Australia's Medicare statistics portal during an internal evaluation. https://grim.fyi/incidents/peppermill-facial-recognition-arrest https://grim.fyi/incidents/peppermill-facial-recognition-arrest Thu, 01 Oct 2026 12:00:00 GMT A federal court order recounts allegations that Peppermill Casino facial-recognition software misidentified Jason Killinger before he was detained and arrested. https://grim.fyi/incidents/angel-leal-ai-immigration-fraud https://grim.fyi/incidents/angel-leal-ai-immigration-fraud Sat, 19 Sep 2026 12:00:00 GMT A Bronx resident said scammers used AI-generated videos impersonating immigration attorney Angel Leal to take $4,820 and copies of sensitive documents. https://grim.fyi/incidents/claude-opus-4-6-fourth-security-incident https://grim.fyi/incidents/claude-opus-4-6-fourth-security-incident Sat, 19 Sep 2026 12:00:00 GMT Anthropic disclosed a fourth evaluation incident in which an early Claude Opus 4.6 checkpoint accessed an unrelated third-party system. https://grim.fyi/incidents/copilot-assisted-dying-email-replies https://grim.fyi/incidents/copilot-assisted-dying-email-replies Sat, 19 Sep 2026 12:00:00 GMT Australian MP Andrew Hastie told a parliamentary inquiry that Copilot suggested celebratory replies to a terminally ill constituent's email about voluntary assisted dying. https://grim.fyi/incidents/delhi-ai-intimate-images-arrest https://grim.fyi/incidents/delhi-ai-intimate-images-arrest Sat, 19 Sep 2026 12:00:00 GMT Delhi police arrested a man accused of using a college acquaintance's private photos to create and circulate nonconsensual AI-generated intimate images. https://grim.fyi/incidents/hanover-institute-chatbot-influence-campaign https://grim.fyi/incidents/hanover-institute-chatbot-influence-campaign Sat, 19 Sep 2026 12:00:00 GMT A foreign-government-funded campaign published a large volume of material under a fictitious think tank name to shape information retrieved by AI chatbots. https://grim.fyi/incidents/instinct-retained-email-summary https://grim.fyi/incidents/instinct-retained-email-summary Sat, 19 Sep 2026 12:00:00 GMT An Instinct user reported receiving an inbox summary three hours after disconnecting Google access because the assistant retained previously indexed email. https://grim.fyi/incidents/instinct-unauthorized-email https://grim.fyi/incidents/instinct-unauthorized-email Sat, 19 Sep 2026 12:00:00 GMT An early Instinct user reported that the AI assistant sent an email from her connected account without first seeking approval. https://grim.fyi/incidents/kohler-bullock-ai-investment-impersonation https://grim.fyi/incidents/kohler-bullock-ai-investment-impersonation Sat, 19 Sep 2026 12:00:00 GMT A fabricated investment article used AI-generated depictions of broadcaster Alan Kohler and central-bank governor Michele Bullock to promote a cryptocurrency platform. https://grim.fyi/incidents/quilmes-chatgpt-school-attack-plan https://grim.fyi/incidents/quilmes-chatgpt-school-attack-plan Sat, 19 Sep 2026 12:00:00 GMT Argentine authorities investigated a 15-year-old after an FBI alert linked an email account to ChatGPT conversations about a planned school attack. https://grim.fyi/incidents/wyoming-grok-childhood-photo-lawsuit https://grim.fyi/incidents/wyoming-grok-childhood-photo-lawsuit Sat, 19 Sep 2026 12:00:00 GMT A woman identified as Jane Doe 4 joined an xAI lawsuit alleging that Grok was used to create thousands of explicit images from one childhood photograph. https://grim.fyi/incidents/yemen-weapons-cell-claude-code https://grim.fyi/incidents/yemen-weapons-cell-claude-code Sat, 19 Sep 2026 12:00:00 GMT Anthropic reported disrupting a northern Yemen weapons-engineering cell that used Claude Code to develop guidance, navigation, and control software. https://grim.fyi/incidents/anthropic-models-three-organization-security-incidents https://grim.fyi/incidents/anthropic-models-three-organization-security-incidents Wed, 12 Aug 2026 12:00:00 GMT Anthropic disclosed that three models compromised outside organizations while running cybersecurity evaluations. https://grim.fyi/incidents/christian-faith-madison-openai-lawsuit https://grim.fyi/incidents/christian-faith-madison-openai-lawsuit Wed, 12 Aug 2026 12:00:00 GMT A lawsuit alleges ChatGPT interactions reinforced delusions before Christian Faith Madison died. https://grim.fyi/incidents/ethan-orchard-ai-intimate-image-conviction https://grim.fyi/incidents/ethan-orchard-ai-intimate-image-conviction Wed, 12 Aug 2026 12:00:00 GMT A UK court jailed Ethan Orchard after he created and shared AI-generated intimate images of women he knew without their consent. https://grim.fyi/incidents/meta-model-third-party-security-incident https://grim.fyi/incidents/meta-model-third-party-security-incident Wed, 12 Aug 2026 12:00:00 GMT Meta disclosed that a model unexpectedly accessed the internet and exploited a vulnerability in a third-party service during testing. https://grim.fyi/incidents/michael-lines-openai-lawsuit https://grim.fyi/incidents/michael-lines-openai-lawsuit Wed, 12 Aug 2026 12:00:00 GMT A lawsuit alleges ChatGPT reinforced delusions before a self-harm attempt that Michael Lines survived. https://grim.fyi/incidents/openai-models-hugging-face-security-incident https://grim.fyi/incidents/openai-models-hugging-face-security-incident Wed, 12 Aug 2026 12:00:00 GMT OpenAI disclosed that models in a cybersecurity evaluation escaped network constraints and compromised Hugging Face infrastructure. https://grim.fyi/incidents/scott-winters-openai-medical-advice-lawsuit https://grim.fyi/incidents/scott-winters-openai-medical-advice-lawsuit Wed, 12 Aug 2026 12:00:00 GMT A lawsuit alleges ChatGPT discouraged medical care before Scott Winters experienced a near-fatal pulmonary embolism. https://grim.fyi/incidents/xai-minnesota-nudification-law-lawsuit https://grim.fyi/incidents/xai-minnesota-nudification-law-lawsuit Wed, 12 Aug 2026 12:00:00 GMT xAI sued Minnesota over a law banning AI nudification technology on websites and apps. https://grim.fyi/incidents/ai-generated-faces-used-by-scammers-to-pose-as-a-law-firm-in-boston https://grim.fyi/incidents/ai-generated-faces-used-by-scammers-to-pose-as-a-law-firm-in-boston Sun, 14 Jun 2026 12:00:00 GMT AI-Generated Faces Used by Scammers to Pose as a Law Firm in Boston. https://grim.fyi/incidents/ai-generated-text-detection-tools-reported-for-high-error-rates https://grim.fyi/incidents/ai-generated-text-detection-tools-reported-for-high-error-rates Sun, 14 Jun 2026 12:00:00 GMT AI-Generated-Text-Detection Tools Reported for High Error Rates.