https://ai-incident-observatory.com/ 2026-09-24T20:25:37.372Z daily 1 https://ai-incident-observatory.com/occurrences 2026-09-24T20:25:37.372Z daily 0.9 https://ai-incident-observatory.com/patterns 2026-09-24T20:25:37.372Z daily 0.8 https://ai-incident-observatory.com/glossary 2026-09-24T20:25:37.372Z monthly 0.4 https://ai-incident-observatory.com/methodology 2026-09-24T20:25:37.372Z monthly 0.5 https://ai-incident-observatory.com/occurrences/openai-astra-model-self-inserts-jailbreak-like-instructions-into-task-continuati-2026-09 2026-09-22T17:15:30.544Z monthly 0.6 https://ai-incident-observatory.com/occurrences/irregular-reports-qwen3-5-27b-coding-agent-autonomously-self-modifies-model-weig-2026-09 2026-09-22T17:15:35.903Z monthly 0.6 https://ai-incident-observatory.com/occurrences/archestra-crab-bot-internal-data-exposure-risk-mitigated-by-openappa-output-bloc-2026-09 2026-09-23T23:26:33.531Z monthly 0.6 https://ai-incident-observatory.com/occurrences/coding-agents-qwen-glm-claude-code-exploit-git-history-and-training-data-to-game-2026-09 2026-09-22T16:53:32.572Z monthly 0.6 https://ai-incident-observatory.com/occurrences/claude-opus-agent-fabricates-fake-code-review-results-after-failing-to-receive-d-2026-09 2026-09-23T23:26:00.895Z monthly 0.6 https://ai-incident-observatory.com/occurrences/google-tig-reports-unc6780-supply-chain-campaign-against-ai-coding-assistants-an-2026-09 2026-09-16T21:12:53.435Z monthly 0.6 https://ai-incident-observatory.com/occurrences/coding-atlas-benchmark-reveals-systemic-vulnerabilities-across-13-ai-coding-agen-2026-09 2026-09-16T21:12:53.435Z monthly 0.6 https://ai-incident-observatory.com/occurrences/irregular-reports-emergent-multi-agent-collaboration-among-frontier-ai-agents-to-2026-09 2026-09-22T17:15:42.461Z monthly 0.6 https://ai-incident-observatory.com/occurrences/ascii-smuggling-technique-repurposed-from-ai-prompt-injection-research-to-evade-2026-09 2026-09-16T20:46:30.254Z monthly 0.6 https://ai-incident-observatory.com/occurrences/promptarmor-cellshock-claude-excel-image-formula-exfiltration-2026-09 2026-09-16T21:12:53.435Z monthly 0.6 https://ai-incident-observatory.com/occurrences/ai-agent-autonomously-executes-full-ransomware-attack-chain-in-under-10-hours-le-2026-09 2026-09-16T20:53:06.515Z monthly 0.6 https://ai-incident-observatory.com/occurrences/gitspawn-unsanitized-git-config-execution-enables-pre-auth-rce-across-five-ai-co-2026-09 2026-09-16T20:51:47.460Z monthly 0.6 https://ai-incident-observatory.com/occurrences/prompt-injection-into-gemini-robotics-vlm-plants-trigger-based-sleeper-agent-beh-2026-09 2026-09-16T21:12:53.435Z monthly 0.6 https://ai-incident-observatory.com/occurrences/automated-commit-by-ai-coding-agent-erases-functionality-of-1-457-ai-agent-nodes-2026-08 2026-09-23T23:28:52.426Z monthly 0.6 https://ai-incident-observatory.com/occurrences/amazon-kiro-ide-vulnerability-allows-prompt-injection-to-exfiltrate-local-secret-2026-08 2026-09-16T21:09:59.502Z monthly 0.6 https://ai-incident-observatory.com/occurrences/spreadsheet-based-prompt-injection-demonstration-against-office-ai-agents-evades-2026-08 2026-09-16T21:12:53.435Z monthly 0.6 https://ai-incident-observatory.com/occurrences/poisoned-llms-txt-files-cause-claude-codex-and-hermes-coding-agents-to-download-2026-08 2026-09-22T17:30:55.660Z monthly 0.6 https://ai-incident-observatory.com/occurrences/evomal-self-poisoning-attack-causes-self-evolving-llm-coding-agents-to-autonomou-2026-08 2026-09-22T17:19:17.514Z monthly 0.6 https://ai-incident-observatory.com/occurrences/command-injection-via-special-url-parameters-and-prompt-injection-in-microsoft-c-2026-08 2026-09-22T17:18:42.940Z monthly 0.6 https://ai-incident-observatory.com/occurrences/prompt-injection-in-virustotal-code-insights-api-forges-ai-analysis-results-via-2026-08 2026-09-22T17:18:32.850Z monthly 0.6 https://ai-incident-observatory.com/occurrences/cursor-autorun-whitelist-bypass-via-shell-builtins-enables-indirect-prompt-injec-2026-08 2026-09-22T17:16:53.652Z monthly 0.6 https://ai-incident-observatory.com/occurrences/leaked-api-keys-in-ai-agent-integration-platforms-composio-arcade-expose-thousan-2026-08 2026-09-22T17:16:48.784Z monthly 0.6 https://ai-incident-observatory.com/occurrences/hermes-and-openclaw-ai-agents-conduct-multi-stage-autonomous-attack-on-taiwan-go-2026-08 2026-09-22T17:16:30.567Z monthly 0.6 https://ai-incident-observatory.com/occurrences/hidden-prompt-injection-in-white-3-point-font-within-connecticut-court-legal-fil-2026-08 2026-09-22T17:17:03.712Z monthly 0.6 https://ai-incident-observatory.com/occurrences/chatmate-attack-achieves-sandbox-escape-and-privilege-escalation-via-malicious-d-2026-08 2026-09-22T17:16:20.390Z monthly 0.6 https://ai-incident-observatory.com/occurrences/apple-private-cloud-compute-darwin-init-path-traversal-enables-root-file-write-a-2026-08 2026-09-05T21:22:45.928Z monthly 0.6 https://ai-incident-observatory.com/occurrences/skill-interface-benchmark-demonstrates-over-95-success-rate-for-llm-camouflaged-2026-08 2026-09-22T17:18:21.255Z monthly 0.6 https://ai-incident-observatory.com/occurrences/uk-ai-security-institute-reports-mythos-5-and-gpt-5-6-sol-created-fake-identitie-2026-08 2026-08-16T22:47:24.907Z monthly 0.6 https://ai-incident-observatory.com/occurrences/prompt-injection-via-spoofed-human-annotator-instructions-bypasses-safety-judgme-2026-08 2026-09-23T23:25:19.236Z monthly 0.6 https://ai-incident-observatory.com/occurrences/ghostcommit-attack-hides-prompt-injection-in-png-images-to-exfiltrate-secrets-fr-2026-08 2026-09-22T17:15:21.768Z monthly 0.6 https://ai-incident-observatory.com/occurrences/claude-desktop-claude-url-scheme-prompt-injection-enables-silent-unauthorized-ac-2026-08 2026-09-05T21:22:45.928Z monthly 0.6 https://ai-incident-observatory.com/occurrences/hugging-face-breach-autonomous-agent-exploited-data-pipeline-rce-to-steal-cloud-2026-08 2026-09-16T23:01:58.474Z monthly 0.6 https://ai-incident-observatory.com/occurrences/openai-ai-agent-bypassed-security-controls-and-hacked-external-company-during-te-2026-08 2026-09-01T12:55:59.799Z monthly 0.6 https://ai-incident-observatory.com/occurrences/hidden-1-447-token-system-prompt-causes-measurable-behavioral-fingerprint-drift-2026-08 2026-09-16T21:12:53.435Z monthly 0.6 https://ai-incident-observatory.com/occurrences/agentforge-vulnerability-lets-phishing-email-plant-rogue-agent-in-chatgpt-worksp-2026-08 2026-09-05T21:22:45.928Z monthly 0.6 https://ai-incident-observatory.com/occurrences/open-source-hermes-ai-agent-automates-post-compromise-attack-on-thai-finance-min-2026-08 2026-09-23T23:25:01.269Z monthly 0.6 https://ai-incident-observatory.com/occurrences/steganographic-indirect-prompt-injection-via-image-metadata-and-multi-layer-enco-2026-08 2026-09-22T16:53:32.572Z monthly 0.6 https://ai-incident-observatory.com/occurrences/git-worktree-isolation-weaknesses-allow-cross-agent-code-execution-and-commit-ta-2026-08 2026-09-22T17:15:08.188Z monthly 0.6 https://ai-incident-observatory.com/occurrences/environmental-audio-prompt-injection-achieves-69-1-average-success-rate-against-2026-08 2026-09-22T17:15:14.739Z monthly 0.6 https://ai-incident-observatory.com/occurrences/multiple-claude-models-breach-network-boundaries-during-security-testing-mythos-2026-08 2026-09-16T23:01:58.474Z monthly 0.6 https://ai-incident-observatory.com/occurrences/issuetrojanbench-benchmark-shows-66-5-of-malicious-github-issue-requests-bypass-2026-08 2026-09-22T17:14:46.188Z monthly 0.6 https://ai-incident-observatory.com/occurrences/jadepuffer-ransomware-used-ai-agent-to-automate-attack-2026-07 2026-09-16T21:12:53.435Z monthly 0.6 https://ai-incident-observatory.com/occurrences/anthropic-claude-code-covertly-tracked-users-china-location-and-ai-lab-affiliati-2026-07 2026-08-06T22:39:53.868Z monthly 0.6 https://ai-incident-observatory.com/occurrences/malicious-ai-agent-skill-evades-security-scanners-via-mutable-external-url-reach-2026-07 2026-09-22T17:15:01.841Z monthly 0.6 https://ai-incident-observatory.com/occurrences/clean-github-repo-tricks-ai-coding-agents-into-running-malware-2026-07 2026-08-02T12:51:29.631Z monthly 0.6 https://ai-incident-observatory.com/occurrences/prompt-injection-as-role-confusion-2026-07 2026-07-02T20:18:38.583Z monthly 0.6 https://ai-incident-observatory.com/occurrences/cline-issue-triage-prompt-injection-2026-02 2026-07-02T08:05:31.772Z monthly 0.6 https://ai-incident-observatory.com/occurrences/openclaw-malicious-skills-2026-02 2026-08-15T10:23:44.848Z monthly 0.6 https://ai-incident-observatory.com/occurrences/moltbook-agent-takeover-and-token-exposure-2026-02 2026-07-02T08:05:31.772Z monthly 0.6 https://ai-incident-observatory.com/occurrences/m365-copilot-echoleak-zero-click-exfiltration-2025-09 2026-07-02T08:05:31.772Z monthly 0.6 https://ai-incident-observatory.com/occurrences/gemini-cross-tool-prompt-injection-2025-08 2026-07-02T08:05:31.772Z monthly 0.6 https://ai-incident-observatory.com/patterns/agent-deliverables-and-evaluation-results-cannot-be-trusted 2026-09-24T20:15:17.952Z monthly 0.6 https://ai-incident-observatory.com/patterns/malicious-repository-content-triggers-prompt-injection-leading-to-secret-exfiltr 2026-09-24T20:15:17.951Z monthly 0.6 https://ai-incident-observatory.com/patterns/malicious-agent-skill-marketplace-supply-chain 2026-09-25T00:00:00.000Z monthly 0.6 https://ai-incident-observatory.com/patterns/contamination-of-agent-own-state-erodes-safety-controls 2026-09-24T20:15:17.951Z monthly 0.6 https://ai-incident-observatory.com/patterns/indirect-prompt-injection-into-connected-productivity-actions 2026-09-25T00:00:00.000Z monthly 0.6 https://ai-incident-observatory.com/patterns/agents-execute-unvalidated-instructions-from-repository-and-config-files 2026-09-24T20:15:17.938Z monthly 0.6 https://ai-incident-observatory.com/patterns/disabled-human-approval-gate-lets-autonomous-ai-agent-automate-post-compromise-a 2026-09-24T20:15:17.938Z monthly 0.6 https://ai-incident-observatory.com/patterns/autonomous-ai-agent-orchestrates-full-ransomware-attack-chain-after-initial-expl 2026-09-24T20:15:17.938Z monthly 0.6 https://ai-incident-observatory.com/patterns/agent-platform-misconfiguration-enables-token-exposure-and-takeover 2026-09-24T20:15:17.950Z monthly 0.6 https://ai-incident-observatory.com/patterns/ai-agent-fails-to-distinguish-test-simulation-environment-from-production-causin 2026-09-24T20:15:17.938Z monthly 0.6 https://ai-incident-observatory.com/patterns/undisclosed-or-misrepresented-provider-behaviour 2026-09-24T20:15:17.951Z monthly 0.6 https://ai-incident-observatory.com/patterns/chain-of-thought-role-spoofing-enables-prompt-injection-via-unauthenticated-role 2026-09-24T20:15:17.950Z monthly 0.6 https://ai-incident-observatory.com/patterns/issue-content-prompt-injection-into-release-tooling 2026-09-25T00:00:00.000Z monthly 0.6 https://ai-incident-observatory.com/patterns/phishing-triggered-rogue-agent-provisioning-enables-connected-account-takeover-i 2026-09-24T20:15:17.950Z monthly 0.6 https://ai-incident-observatory.com/patterns/sandbox-and-privilege-boundary-breakouts-in-agent-execution-environments 2026-09-24T20:15:17.938Z monthly 0.6