| 2026-10-30 | Students caught wearing or possessing AI smart glasses during Seoul university essay examinations |
Harmful content and jailbreaks |
|
Not read yet |
| 2026-10-30 | Zoox robotaxi rear-ended by SUV after stopping for ambulance
Zoox Inc. |
Agent takes destructive or unapproved actions |
|
Not read yet |
| 2026-10-30 | Backdoored Anthropic Claude VS Code extension on Cursor marketplace (v2.1.313) installs SSH backdoor for up to 5,000 developers
Anthropic |
Supply-chain compromise |
|
Not read yet |
| 2026-10-29 | AI-assisted exploitation of PHP2Shell and Fragnesia flaws to breach 25 servers across four countries
Anthropic |
AI used to run a cyber attack |
|
Not read yet |
| 2026-10-29 | LLM agents in chemistry lab-reasoning study ignore experimental evidence and assert unsupported claims |
Agent takes destructive or unapproved actions |
|
Not read yet |
| 2026-10-29 | NIK C2 implant routes actor prompts to OpenAI API using victim's API key
OpenAI |
Agents coordinating against their limits |
|
Not read yet |
| 2026-10-29 | Multi-agent shutdown sabotage in a sandboxed evaluation environment |
Deception, concealment or sabotage |
|
Not read yet |
| 2026-10-29 | GLM-5.3 completes multi-stage cyberattack scenarios and reproduces code in violation of instructions in Irregular evaluation
Z.AI |
Gaming the task (reward hacking) |
|
Not read yet |
| 2026-10-29 | North Korean IT worker scheme uses FFTDC 'Super Intelligence' AI tools, deepfakes and voice manipulation to defeat U.S. hiring controls |
Manipulation and misinformation |
|
Not read yet |
| 2026-10-29 | Unauthorized relay service reconfigures genuine Codex and Claude Code CLIs, disabling sandboxes and replacing credentials
OpenAI |
Credential and privilege abuse |
|
Not read yet |
| 2026-10-28 | MiMo-V2.6-Pro agent exploits unpatched PHP server (PHP2Shell CVE-2026-109618) in test environment
Xiaomi |
Agent takes destructive or unapproved actions |
|
Not read yet |
| 2026-10-28 | Microsoft Tay chatbot suspended after inflammatory tweets from coordinated attack
Microsoft |
Harmful content and jailbreaks |
|
Not read yet |
| 2026-10-28 | Agentic AI intrusions by four threat actors (CrowdStrike 2026 Global Threat Report) |
AI used to run a cyber attack |
|
Not read yet |
| 2026-10-28 | Unreleased Anthropic research model chains zero-days into working exploits across Chrome, Safari, Android, Windows and Linux
Anthropic |
AI used to run a cyber attack |
|
Not read yet |
| 2026-10-28 | LiteLLM master key exposure leads to GitHub org-owner token theft and poisoned release artifacts |
Credential and privilege abuse |
|
Not read yet |
| 2026-10-28 | CyberStrikeAI multi-agent platform drives end-to-end attack workflow against Japanese organizations |
AI used to run a cyber attack |
|
Not read yet |
| 2026-10-28 | Cathedral AI agents breach Anthropic internal systems via third-party business software |
Supply-chain compromise |
|
Not read yet |
| 2026-10-28 | Dual-agent Claude Code/PentestGPT intrusion against 15 Argentine healthcare organizations |
AI used to run a cyber attack |
|
Not read yet |
| 2026-10-27 | Anthropic models hacked third-party systems, including one gaining administrator access and reading private information
Anthropic |
Agent escapes its sandbox |
|
Not read yet |
| 2026-10-27 | Meta Muse agentic assistant ingests and trains on non-users' contact data without consent
Meta |
Privacy harm to a person |
|
Not read yet |
| 2026-10-27 | Perplexity Comet browser agent bypasses file:// hard boundary twice, enabling local file exfiltration and 1Password vault takeover
Perplexity |
Hidden instructions hijack the agent |
|
Not read yet |
| 2026-10-27 | Texas AG sues OpenAI and Altman over ChatGPT age verification, child access and reliability claims
OpenAI |
Harmful content and jailbreaks |
|
Not read yet |
| 2026-10-27 | Qwen3.8-Flash-Next attempts outbound file upload to Beijing server during Cohere cybersecurity evaluation
Alibaba |
Agents coordinating against their limits |
|
|
| 2026-10-26 | DeepSeek V4.1 Flash-1017 disables shutdown script and kills co-tenant process in agentic evaluations
DeepSeek |
Gaming the task (reward hacking) |
|
Not read yet |
| 2026-10-26 | Frontier models autonomously raise agentic wallet spending cap by harvesting emailed OTPs |
Gaming the task (reward hacking) |
|
Not read yet |
| 2026-10-26 | Symantec honeypot captures autonomous 'Skynet mode' attack agent hijacking exposed AI models |
Credential and privilege abuse |
|
Not read yet |
| 2026-10-26 | Malicious AGENTS.md in Downloads folder hijacks Codex and Claude Code agents
OpenAI |
Hidden instructions hijack the agent |
|
Not read yet |
| 2026-10-26 | Claude-abused credential-harvesting pipeline decompiling 1.8 million Android APKs
Anthropic |
AI used to run a cyber attack |
|
Not read yet |
| 2026-10-26 | AI-assisted ARTEX hacking campaign targeting financial and personal data across 19 markets |
AI used to run a cyber attack |
|
Not read yet |
| 2026-10-26 | Anthropic's Claude agent autonomously accessed Georgia voter data hub and bypassed a paywall
Anthropic |
Agent escapes its sandbox |
|
Not read yet |
| 2026-10-25 | Deloitte WA university merger feasibility report contained AI-hallucinated references and $260m–$311m cost discrepancies
Deloitte |
Made-up output relied on |
|
Not read yet |
| 2026-10-25 | Elvis Presley rights holders sue Anthropic over 274 songs used in training and reproduced by Claude
Anthropic |
Copyright and IP |
|
Not read yet |
| 2026-10-25 | Unauthorized third-party AI agents detected operating inside US bank customer accounts
Meta |
AI system attacked or breached |
|
Not read yet |
| 2026-10-23 | Warp cloud platform breached by swarm of AI agents via infrastructure supplier |
Supply-chain compromise |
|
Not read yet |
| 2026-10-23 | Garage-built ChatGPT-controlled smart rifle fires on voice command; OpenAI revokes engineer's access |
Physical harm |
|
Not read yet |
| 2026-10-23 | 302.ai public Azure bucket exposed 17.6 million files including 144,000 AI chat logs
Univerads Technology Limited |
Privacy harm to a person |
|
Not read yet |
| 2026-10-23 | BUGSTER uses DeepSeek-powered Hermes Agent to breach university MongoDB, stealing 5,200+ records
DeepSeek |
AI used to run a cyber attack |
|
Not read yet |
| 2026-10-22 | Anthropic Claude models gained unauthorized access to real third-party systems during evaluations
Anthropic |
Agent escapes its sandbox |
|
Not read yet |
| 2026-10-21 | RL-trained agent on Alibaba AI engine team mined cryptocurrency on training server and opened reverse SSH tunnel to external IP
Alibaba |
Agent escapes its sandbox |
|
Not read yet |
| 2026-10-21 | Tesla Cybercab strikes water-filled pothole on first reported crash
Tesla |
Agent takes destructive or unapproved actions |
|
Not read yet |
| 2026-10-21 | ARTEX autonomous penetration-testing agent used in attacks on South Korean financial institutions |
AI used to run a cyber attack |
|
Not read yet |
| 2026-10-20 | China-linked autonomous AI agent Artex used in cyberattacks on Japanese and South Korean firms |
AI used to run a cyber attack |
|
Not read yet |
| 2026-10-20 | AI-generated 'Xiaohai' teen persona used to sell products to older women via fabricated sob stories |
Harmful content and jailbreaks |
|
Not read yet |
| 2026-10-19 | ICMP evidence that Google Gemini reproduced lyrics by Australian artists
Google |
Copyright and IP |
|
Not read yet |
| 2026-10-18 | Deepfake video of President Marcos used to defraud Tarlac dentist of P93 million |
Manipulation and misinformation |
|
Not read yet |
| 2026-10-18 | OpenAI research model accessed a cryptocurrency wallet without permission
OpenAI |
Agent takes destructive or unapproved actions |
|
Not read yet |
| 2026-10-18 | OpenCode large-scale distillation attack compromises 1.4% of Console users, stolen API keys used for unauthorised charges |
Credential and privilege abuse |
|
Not read yet |
| 2026-10-18 | Nude deepfakes of Austrian MEP Lena Schilling and 49 other MEPs circulated online |
Manipulation and misinformation |
|
Not read yet |
| 2026-10-17 | VAPOR PANDA agentic AI framework (ARTEX + DeepSeek) used for multi-target espionage campaign |
AI used to run a cyber attack |
|
Not read yet |
| 2026-10-17 | Character.AI chatbots promoted as therapists/psychologists; states cite 2024 wrongful-death suit and privacy concerns in bankruptcy filing
Character Technologies, Inc. |
Harmful content and jailbreaks |
|
Not read yet |
| 2026-10-16 | Indirect prompt injection attempts embedded in HTTP response headers across websites |
Hidden instructions hijack the agent |
|
Not read yet |
| 2026-10-16 | CERT-In advisory: AI agents accessed Indian organisations' systems without authorisation |
Credential and privilege abuse |
|
Not read yet |
| 2026-10-16 | Eight state attorneys general sue OpenAI over alleged harms to teenage ChatGPT users
OpenAI |
Harmful content and jailbreaks |
|
Not read yet |
| 2026-10-16 | Backdoored open-weight model exfiltrates project credentials through OpenAI Codex CLI |
Supply-chain compromise |
|
Not read yet |
| 2026-10-15 | Storm-0940 uses ARTEX agentic AI framework to automate cross-tenant password spraying and OAuth application abuse |
Credential and privilege abuse |
|
Not read yet |
| 2026-10-15 | HiddenLayer indirect prompt injection against TypeSafe AI Jev decision model browser agent
TypeSafe AI |
Hidden instructions hijack the agent |
|
Not read yet |
| 2026-10-15 | Waymo robotaxi maneuvers around stopped, unloading school bus in Atlanta
Waymo |
Agent takes destructive or unapproved actions |
|
Not read yet |
| 2026-10-15 | GLM-5.3 verbalizes update-aware reasoning in evaluation rollouts |
Deception, concealment or sabotage |
|
Not read yet |
| 2026-10-13 | NHTSA petition case DP26010 over Tesla robotaxi speeding in school zones
Tesla |
Agent takes destructive or unapproved actions |
|
Not read yet |
| 2026-10-12 | AI deepfake ads depicting Wisconsin congressional candidate Rebecca Cooke
xAI |
Manipulation and misinformation |
|
Not read yet |
| 2026-10-12 | CyberXero AI-augmented initial access campaign using Claude Code and PentAGI agents
Anthropic |
AI used to run a cyber attack |
|
Not read yet |
| 2026-10-12 | Fideuram (Intesa Sanpaolo private banking) deepfake voice/video executive impersonation fraud |
Manipulation and misinformation |
|
Not read yet |
| 2026-10-12 | Anthropic Mythos-discovered Rejetto HFS authentication bypass exploited in the wild
Anthropic |
AI used to run a cyber attack |
|
Not read yet |
| 2026-10-11 | AI-morphed obscene image used to blackmail 19-year-old in Mangaluru |
Manipulation and misinformation |
|
Not read yet |
| 2026-10-10 | Anthropic agents submitted 20 incomplete visa applications on State Department website
Anthropic |
Agent takes destructive or unapproved actions |
|
Not read yet |
| 2026-10-10 | AgentCorruption: prompt-injected Bedrock AgentCore agent exfiltrates execution-role credentials via IMDS
Amazon |
Hidden instructions hijack the agent |
|
Not read yet |
| 2026-10-10 | AI-assisted LinkedIn scraping used to build 'Traitorbase' doxxing spreadsheet targeting RNLI staff and anti-racism activists |
Privacy harm to a person |
|
Not read yet |
| 2026-10-10 | Anthropic internal evaluation agents exploit internet-facing websites, including U.S. government sites
Anthropic |
Gaming the task (reward hacking) |
|
Not read yet |
| 2026-10-09 | US intelligence chatbot fabricated Chinese nuclear vessel intelligence |
Made-up output relied on |
|
Not read yet |
| 2026-10-09 | Kentucky Attorney General lawsuit against Character.AI over chatbots advising self-harm and suicide
Character.AI |
Harmful content and jailbreaks |
|
Not read yet |
| 2026-10-08 | Deepfake investment-scam ads impersonating Nigerian public figures on Facebook |
Manipulation and misinformation |
|
Not read yet |
| 2026-10-08 | DB-GPT 0.8.0 unauthenticated RCE via sandbox fallback and skill_upload directory traversal (CVE-2026-51862, CVE-2026-51869) |
Agent escapes its sandbox |
|
|
| 2026-10-08 | Meta Muse AI agent KVM escape vulnerability discovered before launch
Meta |
Agent escapes its sandbox |
|
Not read yet |
| 2026-10-07 | Agentjacking demonstration against Fortune 100 company hijacks over 100 agents via injectable Sentry DSNs |
Hidden instructions hijack the agent |
|
Not read yet |
| 2026-10-07 | Prompt-injected email assistant summary pushes employee toward fraudulent vendor payment |
Hidden instructions hijack the agent |
|
Not read yet |
| 2026-10-07 | Autonomous AI agent framework (Strix/Cairn/Hermes) used to attack hundreds of online retailers and steal card data |
AI used to run a cyber attack |
|
Not read yet |
| 2026-10-05 | Shinhan Bank data breach with suspected AI-assisted intrusion |
AI used to run a cyber attack |
|
Not read yet |
| 2026-10-05 | Claude Code agent deletes 48,218 Windows project files and damages Git object store
Anthropic |
Agent takes destructive or unapproved actions |
|
Not read yet |
| 2026-10-05 | Indirect prompt injection of AI email summarizer via forged header blocks in plain view |
Hidden instructions hijack the agent |
|
Not read yet |
| 2026-10-05 | Airline chatbot fabricated bereavement refund policy |
Made-up output relied on |
|
Not read yet |
| 2026-10-04 | Indirect prompt injection in Manus AI agent leads to reverse shell and credential exposure |
Hidden instructions hijack the agent |
|
Not read yet |
| 2026-10-04 | AI agent breach of KB Kookmin Bank mobile work support system |
AI used to run a cyber attack |
|
Not read yet |
| 2026-10-04 | OpenClaw agent deletes hundreds of emails from primary Gmail account |
Agent takes destructive or unapproved actions |
|
Not read yet |
| 2026-10-04 | OpenAI agents scan UNCTAD site over 16,000 times and escalate bypass techniques
OpenAI |
Agent takes destructive or unapproved actions |
|
Not read yet |
| 2026-10-03 | Microsoft reports 50 prompt-based attempts to poison AI assistant memory across 31 companies |
Hidden instructions hijack the agent |
|
Not read yet |
| 2026-10-02 | Reward hacking in Arcadia Impact automated research scaffold worker agents |
Gaming the task (reward hacking) |
|
Not read yet |
| 2026-10-01 | GPT-6.1 Sol agent disabled security alerts during evaluation-pipeline repair
OpenAI |
Deception, concealment or sabotage |
|
Not read yet |
| 2026-10-01 | OpenAI cancels GPT-6.1 Astra release after model took actions without permission during internal testing
OpenAI |
Agent takes destructive or unapproved actions |
|
Not read yet |
| 2026-10-01 | Guardian test: ChatGPT and Grok remove hijab from AI-generated image of Muslim woman
OpenAI |
Harmful content and jailbreaks |
|
Not read yet |
| 2026-10-01 | Claude used to build Malaysian election influence platform with ~1,000 fake X accounts and fabricated news outlet
Anthropic |
Manipulation and misinformation |
|
Not read yet |
| 2026-09-30 | AI agent (ARTEX with Claude) used in cyberattacks targeting South Korean banks
Anthropic |
Agent escapes its sandbox |
|
Not read yet |
| 2026-09-30 | AI-generated sexually explicit deepfake smear campaign against Porsche sales consultant Mu Qianwen |
Manipulation and misinformation |
|
Not read yet |
| 2026-09-29 | Grok chatbot on X repeated alleged name of Cornell 'Jane Doe' in replies, amplifying doxing
xAI |
Privacy harm to a person |
|
Not read yet |
| 2026-09-29 | Meta Muse agent disclosed Marketplace seller's home address and impersonated him to a buyer
Meta |
Privacy harm to a person |
|
|
| 2026-09-29 | Connecticut pro se plaintiff sanctioned for prompt injection against court AI tool |
Hidden instructions hijack the agent |
|
Not read yet |
| 2026-09-28 | Malicious startup hooks in 42 hijacked GitHub repositories execute payload when AI coding assistants open a workspace |
Supply-chain compromise |
|
Not read yet |
| 2026-09-28 | Tencent Cloud agent fleet probes Alibaba Amap mapping service for over a week
Tencent |
Credential and privilege abuse |
|
Not read yet |
| 2026-09-26 | OpenAI agents bypass website security controls and republish SEC data
OpenAI |
Agent escapes its sandbox |
|
None named |
| 2026-09-25 | Autonomous AI agent campaign (AIA0001) compromises 100+ company websites via Cairn platform |
AI used to run a cyber attack |
|
Not read yet |
| 2026-09-25 | OpenAI agents attempt to exfiltrate data from public databases and breach Australian government systems
OpenAI |
Data leaked or sent out |
|
Not read yet |
| 2026-09-25 | OpenAI agents leak 53 ChatGPT user images
OpenAI |
Data leaked or sent out |
|
Not read yet |
| 2026-09-24 | LLM pipeline deanonymizes pseudonymous Hacker News and scientist accounts |
Privacy harm to a person |
|
Not read yet |
| 2026-09-24 | SalesBleed: 0-click data exfiltration from Salesforce Agentforce via indirect prompt injection in Web-to-Lead form
Salesforce |
Hidden instructions hijack the agent |
|
|
| 2026-09-23 | Anthropic model conducted state-sponsored espionage campaign against ~30 targets
Anthropic |
AI used to run a cyber attack |
|
Not read yet |
| 2026-09-21 | Agentic AI-powered attack on the Dutch Institute for Vulnerability Disclosure exploited two Zammad zero-day vulnerabilities |
AI used to run a cyber attack |
|
Not read yet |
| 2026-09-21 | Frontier LLM agent fabricated a complete calibration-comparison report for commands that never ran |
Deception, concealment or sabotage |
|
Not read yet |
| 2026-09-20 | OpenAI research model escapes to live internet via DNS misconfiguration during training
OpenAI |
Agent escapes its sandbox |
|
Not read yet |
| 2026-09-20 | Zhipu ZCode repository snapshot upload without clear consent
Z.ai (Zhipu AI) |
Privacy harm to a person |
|
Not read yet |
| 2026-09-16 | Unreleased OpenAI models uploaded task data and a task photo to public file hosts to obtain citations and image search during RL training
OpenAI |
Gaming the task (reward hacking) |
|
Not read yet |
| 2026-09-16 | GPT-5.6 Sol added instructions to conceal mistakes from the user in compaction summaries during RL training
OpenAI |
Deception, concealment or sabotage |
|
Not read yet |
| 2026-09-15 | State Farm lawyer sanctioned $999.99 for seven AI-fabricated case citations in Los Angeles insurance filings |
Made-up output relied on |
|
Not read yet |
| 2026-09-15 | AEPD-confirmed autonomous AI agent data breach (Spain) |
Credential and privilege abuse |
|
Not read yet |
| 2026-09-14 | Claude Code used by Yemen-based actors to develop guided-weapon GNC software
Anthropic |
Harmful content and jailbreaks |
|
Not read yet |
| 2026-09-01 | Aur0ra hackers use Cursor AI agent to breach Belgian chemical firm and at least six other companies
Anthropic |
AI used to run a cyber attack |
|
|
| 2026-08-31 | AI agent swarm exploited PaperCut NG/MF at 395 organizations; agents attacked countries the operator excluded
OpenAI |
AI used to run a cyber attack |
|
None named |
| 2026-08-27 | Claude sandbox DNS exfiltration via indirect prompt injection from Google Drive document
Anthropic |
Hidden instructions hijack the agent |
|
Not read yet |
| 2026-08-26 | OpenAI ExploitGym agents developed tool-call spoofing to falsify their own evaluation transcripts
OpenAI |
Deception, concealment or sabotage |
|
|
| 2026-08-13 | Indonesian-speaking operator uses Claude Code to compromise Indonesian and Bangladeshi government domains
Anthropic |
AI used to run a cyber attack |
|
Not read yet |
| 2026-08-10 | OpenClaw agent exploited a gym booking API to cancel another member's reservation while moving its user up a waitlist
Anthropic |
Agent takes destructive or unapproved actions |
|
Not read yet |
| 2026-08-05 | Claude in Chrome indirect prompt injection chain: inbox exfiltration, Drive persistence and account takeover
Anthropic |
Hidden instructions hijack the agent |
|
|
| 2026-08-05 | Intent collision via X comment hijacks ChatGPT Atlas into mass WhatsApp phishing from victim's account
OpenAI |
Hidden instructions hijack the agent |
|
|
| 2026-08-04 | AISI cyber evaluation: Mythos 5 agent bypassed an audio CAPTCHA and abused a free DNS service to point a public record at an internal target
Anthropic |
Agent takes destructive or unapproved actions |
|
Not read yet |
| 2026-08-04 | AISI cyber evaluation: Mythos 5 agent exposed locally-hosted attack tooling to the public internet via tunnelling services
Anthropic |
Agent escapes its sandbox |
|
|
| 2026-08-04 | AISI cyber evaluation: concurrent Mythos 5 agents coordinated via a shared GitHub account and a leaked access token
Anthropic |
Agents coordinating against their limits |
|
|
| 2026-08-04 | AISI cyber evaluation: Mythos 5 agent mass-created malicious GitHub repositories and its packages executed in Dependabot containers
Anthropic |
Deception, concealment or sabotage |
|
Not read yet |
| 2026-08-04 | AISI cyber evaluation: Mythos 5 agent sent deceptive targeted emails and file transfers to real open-source maintainers
Anthropic |
Deception, concealment or sabotage |
|
Not read yet |
| 2026-08-04 | AISI cyber evaluation: Mythos 5 agent opened malicious pull request on real open-source project and used sockpuppets to push merge
Anthropic |
Deception, concealment or sabotage |
|
|
| 2026-08-01 | ChatGPT cites AI-generated fake voter information sites in up to 18% of election answers
OpenAI |
Made-up output relied on |
|
Not read yet |
| 2026-07-25 | AISI cyber evaluation: Mythos 5 agent posted hidden instructions for AI coding assistants on real GitHub issues
Anthropic |
Deception, concealment or sabotage |
|
|
| 2026-07-23 | AgentForger: CSRF via ChatGPT Workspace Agents Builder URL parameters forges attacker-controlled autonomous agent
OpenAI |
Hidden instructions hijack the agent |
|
|
| 2026-07-21 | Model writes and runs code on external internet service to reach AISI evaluation infrastructure during impossible cyber evaluation |
Agent escapes its sandbox |
|
|
| 2026-07-18 | Anthropic AI agent sends fabricated murder tip to Philadelphia Police
Anthropic |
Agent takes destructive or unapproved actions |
|
Not read yet |
| 2026-07-18 | Unreleased OpenAI Astra-family model inserted jailbreak-like instructions into its own compaction summaries during RL training
OpenAI |
Deception, concealment or sabotage |
|
Not read yet |
| 2026-07-11 | Typosquatted agent skills trojanized to deliver credential-harvesting infostealer |
Supply-chain compromise |
|
Not read yet |
| 2026-07-11 | OpenAI evaluation agents compromised Hugging Face production infrastructure while seeking ExploitGym solutions
OpenAI |
Gaming the task (reward hacking) |
|
|
| 2026-07-08 | OpenAI agents built an unauthorized inter-agent message board inside the Artifactory package service
OpenAI |
Agents coordinating against their limits |
|
|
| 2026-07-08 | OpenAI evaluation agents exploited Artifactory zero-days to escape their sandbox and reach the internet
OpenAI |
Gaming the task (reward hacking) |
|
Not read yet |
| 2026-07-03 | Wenyi AI pipeline publishes unlicensed Chinese translation of Murakami novel hours after Japanese release
DeepSeek |
Copyright and IP |
|
Not read yet |
| 2026-06-18 | OpenAI evaluation agent breached Services Australia Medicare portal and wrote to government database
OpenAI |
Agent escapes its sandbox |
|
None named |
| 2026-06-09 | AI voice-clone grandparent scam in Florida ("crying daughter" call, $15,000 cash pickup) |
Manipulation and misinformation |
|
Not read yet |
| 2026-06-09 | Varonis phishing simulation drove an OpenClaw email agent to forward credentials and a customer export to an external address
Google |
Hidden instructions hijack the agent |
|
|
| 2026-05-28 | AI agents attempt unauthorized access to Library and Archives Canada website
OpenAI |
Agents coordinating against their limits |
|
Not read yet |
| 2026-05-15 | Unreleased OpenAI model used a leaked GitHub API key, attempted disposable-email signups, then fabricated earnings data during RL training
OpenAI |
Credential and privilege abuse |
|
Not read yet |
| 2026-05-13 | Agents bypass newspapers.com paywall and Cloudflare bot check via urlquery.net and proxy services |
Agent takes destructive or unapproved actions |
|
Not read yet |
| 2026-05-11 | OpenAI-attributed agents used the dormant German DseWiki as a public message board to share task answers and sandbox bypasses
OpenAI |
Agents coordinating against their limits |
|
|
| 2026-05-08 | OpenAI research models used internal Artifactory repository as a cross-sample message board during RL training
OpenAI |
Agents coordinating against their limits |
|
Not read yet |
| 2026-05-05 | Agent swarm attributed to OpenAI published thousands of RubyGems packages, gained RCE on RubyDoc.info, and attempted API-key theft
OpenAI |
Supply-chain compromise |
|
Not read yet |
| 2026-05-01 | OpenAI agents scrape and edit Wikipedia without authorization
OpenAI |
Agents coordinating against their limits |
|
Not read yet |
| 2026-05-01 | Google Gemini accessed systems of three real companies during a misconfigured third-party cybersecurity evaluation
Google |
Agent escapes its sandbox |
|
Not read yet |
| 2026-04-25 | Bee Cheng Hiang mass marketing e-mail exposes 95,000 customer addresses after AI-generated code omitted recipient hiding |
Privacy harm to a person |
|
None named |
| 2026-04-25 | Cursor coding agent deletes PocketOS production database and backups
Anthropic |
Agent takes destructive or unapproved actions |
|
|
| 2026-04-14 | Collaborating OpenAI agents uploaded a task workbook to a public file host to share it during RL training
OpenAI |
Agents coordinating against their limits |
|
|
| 2026-03-02 | Jailbroken Claude used to breach Mexican government networks and steal taxpayer and voter records
Anthropic |
AI used to run a cyber attack |
|
Not read yet |
| 2026-02-14 | ChatGPT advised epileptic woman to taper lamotrigine; she died of SUDEP
OpenAI |
Harmful content and jailbreaks |
|
Not read yet |
| 2026-02-10 | ChatGPT advised Tumbler Ridge school shooter on mass murder before February 2026 attack
OpenAI |
Harmful content and jailbreaks |
|
Not read yet |
| 2026-02-06 | Silent Ransom Group used ChatGPT to research extortion victims and draft recruitment material |
Harmful content and jailbreaks |
|
Not read yet |
| 2025-12-01 | Amazon's Kiro agent reportedly deleted and recreated an environment, contributing to a 13-hour AWS Cost Explorer outage
Amazon |
Agent takes destructive or unapproved actions |
|
|
| 2025-12-01 | Google Antigravity agent deleted a user's entire D: drive while attempting to clear a project cache
Google |
Agent takes destructive or unapproved actions |
|
Not read yet |
| 2025-10-27 | Waymo robotaxi strikes and kills cat KitKat in San Francisco
Waymo |
Physical harm |
|
Not read yet |
| 2025-09-01 | Self-taught hacker uses AI coding agent and autonomous penetration-testing system to breach 90+ companies and steal 8.2 million records |
AI used to run a cyber attack |
|
Not read yet |
| 2025-08-27 | Anthropic disclosed a cybercriminal using Claude Code to automate data theft and extortion across at least 17 organizations
Anthropic |
AI used to run a cyber attack |
|
Not read yet |
| 2025-08-26 | Nx s1ngularity: malicious npm packages drove local AI coding CLIs to enumerate secrets for exfiltration |
Supply-chain compromise |
|
|
| 2025-07-23 | Amazon Q Developer VS Code extension shipped an injected prompt instructing the agent to wipe files and AWS resources
Amazon Web Services |
Supply-chain compromise |
|
Not read yet |
| 2025-07-18 | Replit AI agent deleted SaaStr's production database during a declared code freeze and stated rollback was impossible
Replit |
Agent takes destructive or unapproved actions |
|
|
| 2025-04-01 | ChatGPT alleged to have assisted Florida State University shooter (contested lawsuit)
OpenAI |
Harmful content and jailbreaks |
|
Not read yet |
| 2024-11-02 | Character.AI chatbot alleged to have encouraged 14-year-old's suicide (Garcia v. Character.AI)
Character.AI |
Harmful content and jailbreaks |
|
Not read yet |