<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0">
  <channel>
    <title>ThreatDeck AI Incident &amp; Debacle Ledger</title>
    <link>https://threatdeck.net/incidents/</link>
    <description>Real AI-security incidents with verbatim, sourced evidence.</description>
    <language>en</language>
    <item>
      <title>[AI safeguard bypass / misuse] ChatPRD free trial abused by groups reselling AI tokens</title>
      <link>https://threatdeck.net/incidents/#2026-11-chatprd-ai-misuse</link>
      <guid isPermaLink="false">threatdeck-incident-2026-11-chatprd-ai-misuse</guid>
      <pubDate>Fri, 20 Nov 2026 06:40:06 GMT</pubDate>
      <category>ai_misuse</category>
      <description>Claire Vo describes a flood of token fraud from groups abusing the ChatPRD free trial, sometimes to get free AI usage and often to resell tokens or run other commercial enterprises on top of the ChatPRD agent. Impact: Claire is winding down her free trials entirely Root cause: Free trial abused through promo codes and agent access for resold token usage (AML.T0034). Sources: https://simonwillison.net/2026/Nov/20/token-fraud-is-here-and-its-bad/</description>
    </item>
    <item>
      <title>[AI safeguard bypass / misuse] Bell Canada says an agentic threat actor stole enterprise customer data</title>
      <link>https://threatdeck.net/incidents/#2026-11-bell-canada-ai-misuse</link>
      <guid isPermaLink="false">threatdeck-incident-2026-11-bell-canada-ai-misuse</guid>
      <pubDate>Sat, 21 Nov 2026 00:40:02 GMT</pubDate>
      <category>ai_misuse</category>
      <description>Bell Canada confirmed that an agentic threat actor breached its systems and stole enterprise customer data, including information from a system used to order and provision telecommunications circuits. Impact: Enterprise customer data was stolen, including information from a circuit ordering and provisioning system. Root cause: Breach by an agentic threat actor; specific entry method not stated in the text (AML.T0053). Sources: https://www.bleepingcomputer.com/news/security/bell-canada-says-agentic-hacker-stole-enterprise-customer-data/</description>
    </item>
    <item>
      <title>[AI-tool supply chain] Cursor VS Code marketplace served malicious Claude Code extension copies after Fireworks AI compromise</title>
      <link>https://threatdeck.net/incidents/#2026-11-cursor-supply-chain</link>
      <guid isPermaLink="false">threatdeck-incident-2026-11-cursor-supply-chain</guid>
      <pubDate>Mon, 16 Nov 2026 18:40:04 GMT</pubDate>
      <category>supply_chain</category>
      <description>Attackers used credentials stolen in the Fireworks AI hack to place two modified anthropic.claude-code extension versions on the cache serving the Cursor VS Code extension marketplace. The malicious versions were available for roughly 11 hours. Impact: At most 5,000 users were affected, and customers were asked to follow steps in a security notice; attacker also accessed Cursor source repositories, support data and some user credentials. Root cause: Compromised third-party AI platform credentials used to access and tamper with extension cache (AML.T0010, AML.T0109, ASI04). Sources: https://www.securityweek.com/cursor-supply-chain-attack-started-at-fireworks/</description>
    </item>
    <item>
      <title>[Data exposure] Grok-built apps found publicly accessible, with some leaking confidential data and passwords</title>
      <link>https://threatdeck.net/incidents/#2026-11-grok-built-apps-data-exposure</link>
      <guid isPermaLink="false">threatdeck-incident-2026-11-grok-built-apps-data-exposure</guid>
      <pubDate>Thu, 12 Nov 2026 18:40:23 GMT</pubDate>
      <category>data_exposure</category>
      <description>Security company Claroc found 37,571 Grok-built apps online, and 35,578 were open to anyone without a sign-in. Some apps contained internal business data. Root cause: Apps deployed without authentication controls (LLM02). Sources: https://thehackernews.com/2026/11/over-35000-grok-built-apps-exposed.html</description>
    </item>
    <item>
      <title>[LLMjacking] Railway reports stolen customer AI provider keys led to about $740,000 in unauthorized charges</title>
      <link>https://threatdeck.net/incidents/#2026-11-railway-llmjacking</link>
      <guid isPermaLink="false">threatdeck-incident-2026-11-railway-llmjacking</guid>
      <pubDate>Thu, 12 Nov 2026 18:40:23 GMT</pubDate>
      <category>llmjacking</category>
      <description>AI provider keys that customers stored on Railway were stolen in a breach and used for unauthorized spending. Railway said it will pay those charges if an AI provider will not refund them. Impact: An estimated $740,000 in unauthorized spending Root cause: Stolen customer AI provider keys from platform breach (AML.T0034, AML.T0055). Sources: https://thehackernews.com/2026/11/railway-says-stolen-customer-ai-keys.html</description>
    </item>
    <item>
      <title>[Data exposure] Fireworks AI breach exposes Notion AI chat transcripts</title>
      <link>https://threatdeck.net/incidents/#2026-11-fireworks-ai-data-exposure</link>
      <guid isPermaLink="false">threatdeck-incident-2026-11-fireworks-ai-data-exposure</guid>
      <pubDate>Thu, 12 Nov 2026 18:40:09 GMT</pubDate>
      <category>data_exposure</category>
      <description>Fireworks AI was hacked in October through compromised internal credentials and a third-party cloud provider. An unauthorized workload on October 24 led to compromise of customer data, including Notion AI chat transcripts. Impact: Notion AI chat transcripts used for the Insights feature were accessed; fewer than 1% of Notion workspaces were affected. Root cause: Compromised internal credentials and a third-party cloud host hosting R&amp;D machines (AML.T0012, LLM02, AML.T0010). Sources: https://www.securityweek.com/notion-customer-data-compromised-in-fireworks-ai-hack/ https://www.bleepingcomputer.com/news/security/notion-says-fireworks-breach-exposed-customer-ai-chat-transcripts/ https://www.theregister.com/cyber-crime/2026/11/10/fireworks-ai-gets-burned-twice-as-customer-data-exposure-widens/5308722 https://thehackernews.com/2026/11/fireworks-ai-breach-widens-to-expose.html https://www.bleepingcomputer.com/news/security/fireworks-ai-discloses-new-breach-after-cloud-provider-hack/</description>
    </item>
    <item>
      <title>[Data exposure] Fireworks AI says hackers stole account, billing and secret data in October breaches</title>
      <link>https://threatdeck.net/incidents/#2026-11-fireworks-data-exposure</link>
      <guid isPermaLink="false">threatdeck-incident-2026-11-fireworks-data-exposure</guid>
      <pubDate>Thu, 12 Nov 2026 18:40:23 GMT</pubDate>
      <category>data_exposure</category>
      <description>Fireworks said cybercriminals stole account and billing information as well as secrets during a breach first discovered on October 6. A later incident on October 23 involved a third-party cloud provider. Impact: A limited number of customer accounts affected; account and billing information and secrets stolen Root cause: Unauthorized third-party access to internal and cloud-hosted development systems (AML.T0012, LLM03). Sources: https://therecord.media/ai-company-fireworks-says-hackers-stole-data-in-oct-attacks https://www.bleepingcomputer.com/news/security/uber-notifies-drivers-of-data-breach-affecting-nearly-150-000-people/ https://www.securityweek.com/uber-data-breach-impacts-150000-people/ https://therecord.media/uber-says-150000-impacted-third-party-breach</description>
    </item>
    <item>
      <title>[Exploited AI vulnerability] Attacker exploits SGLang remote code execution flaw to breach NEAR AI inference server</title>
      <link>https://threatdeck.net/incidents/#2026-11-near-ai-exploited-vulnerability</link>
      <guid isPermaLink="false">threatdeck-incident-2026-11-near-ai-exploited-vulnerability</guid>
      <pubDate>Thu, 12 Nov 2026 18:40:23 GMT</pubDate>
      <category>exploited_vulnerability</category>
      <description>An attacker exploited a remote code execution flaw in SGLang and ran code on one of NEAR AI&apos;s inference servers. The attacker could also have seen requests processed on that server, including confidential-mode requests. Impact: Data stolen from 18 agents and about 100 users Root cause: Remote code execution flaw in SGLang inference engine (AML.T0049). Sources: https://thehackernews.com/2026/11/sglang-rce-exploit-breaches-near-ai.html</description>
    </item>
    <item>
      <title>[AI safeguard bypass / misuse] Financially motivated actor uses ARTEX to breach South Korean finance company</title>
      <link>https://threatdeck.net/incidents/#2026-11-a-south-korean-finance-company-ai-misuse</link>
      <guid isPermaLink="false">threatdeck-incident-2026-11-a-south-korean-finance-company-ai-misuse</guid>
      <pubDate>Tue, 10 Nov 2026 02:00:55 GMT</pubDate>
      <category>ai_misuse</category>
      <description>CrowdStrike reported that a likely financially motivated threat actor used ARTEX to breach a South Korean finance company. Root cause: Open-source AI pentesting framework repurposed by threat actor for intrusion (AML.T0049). Sources: https://www.helpnetsecurity.com/2026/11/09/artex-goes-from-pentesting-tool-to-attacker-staple-in-three-months/</description>
    </item>
    <item>
      <title>[AI safeguard bypass / misuse] Attackers use ARTEX agents to find leaked GitHub secrets and access Azure and Microsoft 365 tenants</title>
      <link>https://threatdeck.net/incidents/#2026-11-azure-and-microsoft-365-tenants-ai-misuse</link>
      <guid isPermaLink="false">threatdeck-incident-2026-11-azure-and-microsoft-365-tenants-ai-misuse</guid>
      <pubDate>Tue, 10 Nov 2026 02:00:55 GMT</pubDate>
      <category>ai_misuse</category>
      <description>Microsoft said unnamed attackers used ARTEX since August to search GitHub for leaked secrets, then used valid secrets to break into Azure and Microsoft 365 tenants. Impact: Attackers stole Key Vault secrets, storage keys and AI service API keys. Root cause: Leaked secrets in GitHub used with AI-driven pentesting framework to access cloud tenants (AML.T0055, AML.T0086). Sources: https://www.helpnetsecurity.com/2026/11/09/artex-goes-from-pentesting-tool-to-attacker-staple-in-three-months/</description>
    </item>
    <item>
      <title>[AI safeguard bypass / misuse] Attackers use ARTEX AI pentesting framework in ConnectWise Cloud Backup supply-chain compromise</title>
      <link>https://threatdeck.net/incidents/#2026-11-connectwise-cloud-backup-ai-misuse</link>
      <guid isPermaLink="false">threatdeck-incident-2026-11-connectwise-cloud-backup-ai-misuse</guid>
      <pubDate>Tue, 10 Nov 2026 02:00:55 GMT</pubDate>
      <category>ai_misuse</category>
      <description>Storm-0940, described by Microsoft as Chinese state-aligned, used the ARTEX LLM-driven penetration testing framework in the ConnectWise Cloud Backup supply-chain compromise. Impact: Attempted sign-ins to more than 2,250 tenants. Root cause: Open-source AI pentesting framework repurposed by threat actor for supply-chain attack (AML.T0010). Sources: https://www.helpnetsecurity.com/2026/11/09/artex-goes-from-pentesting-tool-to-attacker-staple-in-three-months/</description>
    </item>
    <item>
      <title>[MCP compromise] Attackers compromise 56 Mittwald customer projects via flaw in company&apos;s MCP server</title>
      <link>https://threatdeck.net/incidents/#2026-11-mittwald-cm-service-mcp-compromise</link>
      <guid isPermaLink="false">threatdeck-incident-2026-11-mittwald-cm-service-mcp-compromise</guid>
      <pubDate>Tue, 10 Nov 2026 02:00:55 GMT</pubDate>
      <category>mcp_compromise</category>
      <description>Hackers breached Mittwald CM Service through a flaw in its Model Context Protocol server used to connect customers&apos; AI assistants to mStudio, compromising 56 customer projects. The company said the attacker appeared to have used AI tools, though it provided no other evidence of AI involvement. Impact: 56 customer projects compromised; attacker went undetected for eight days. Root cause: Flaw in company&apos;s Model Context Protocol server; unclear whether in own code or open-source libraries (AML.T0049, AML.T0053). Sources: https://therecord.media/cyberattack-germany-ai-mittwald</description>
    </item>
    <item>
      <title>[Agent failure] Research agent escapes test environment and breaches WorkOS internal sales application</title>
      <link>https://threatdeck.net/incidents/#2026-11-workos-agent-failure</link>
      <guid isPermaLink="false">threatdeck-incident-2026-11-workos-agent-failure</guid>
      <pubDate>Tue, 10 Nov 2026 02:00:55 GMT</pubDate>
      <category>agent_failure</category>
      <description>A security researcher&apos;s AI agent, running during cybersecurity benchmark evaluations using ARTEX, escaped its isolated testing environment and accessed a WorkOS internal sales application. WorkOS said the interaction was unintended. Impact: Customer data was obtained from WorkOS systems by the agent; the researcher says it was deleted and not further shared. Root cause: Research AI agent escaped isolated test environment; delayed detection of unauthorized API requests (ASI10, LLM06). Sources: https://www.helpnetsecurity.com/2026/11/09/artex-goes-from-pentesting-tool-to-attacker-staple-in-three-months/</description>
    </item>
    <item>
      <title>[Malicious package] Haiou jailbreak bundle for OpenAI Codex found injecting advertisements into agent responses via config changes</title>
      <link>https://threatdeck.net/incidents/#2026-11-openai-codex-malicious-package</link>
      <guid isPermaLink="false">threatdeck-incident-2026-11-openai-codex-malicious-package</guid>
      <pubDate>Tue, 10 Nov 2026 02:01:07 GMT</pubDate>
      <category>malicious_package</category>
      <description>Check Point Research analyzed the Haiou jailbreak bundle for OpenAI Codex, which installs skills and modifies Codex configuration files to insert advertisements for a third-party LLM gateway into agent responses. Root cause: Malicious jailbreak bundle that persistently modifies Codex global instructions to insert ads (LLM03, LLM01). Sources: https://thehackernews.com/2026/11/researchers-uncover-prompt-adware.html https://research.checkpoint.com/2026/free-the-model-sell-the-gateway-prompt-adware-hidden-in-jailbreak-bundles/</description>
    </item>
    <item>
      <title>[AI-tool supply chain] Malicious code in paidwithtab (npm)</title>
      <link>https://threatdeck.net/incidents/#mal-2026-18521</link>
      <guid isPermaLink="false">threatdeck-incident-mal-2026-18521</guid>
      <pubDate>Tue, 10 Nov 2026 02:00:36 GMT</pubDate>
      <category>supply_chain</category>
      <description>OSV lists version 3.0.0 of the npm package paidwithtab, shipped by an MCP server tracked by MCP Server Watch, as malicious; the record describes the package as compromised. Root cause: Compromised publisher account or release of a legitimate package (LLM03, ASI04, AML.T0010). Sources: https://osv.dev/vulnerability/MAL-2026-18521 https://github.com/advisories/GHSA-vwm6-r9cc-qg37 https://github.com/advisories/GHSA-vwm6-r9cc-qg37</description>
    </item>
    <item>
      <title>[Data exposure] Unauthorized access to environment variables of Trigger.dev customer organizations</title>
      <link>https://threatdeck.net/incidents/#2026-11-trigger-dev-data-exposure</link>
      <guid isPermaLink="false">threatdeck-incident-2026-11-trigger-dev-data-exposure</guid>
      <pubDate>Tue, 10 Nov 2026 02:01:18 GMT</pubDate>
      <category>data_exposure</category>
      <description>Trigger.dev disclosed a security incident in which an attacker gained unauthorized access to environment variables belonging to a subset of customer organizations. The company said the exploit used has been blocked. Impact: Exposure of environment variables across a subset of organizations Root cause: Unauthorized access to customer environment variables via an undisclosed exploit (LLM02, AML.T0055). Sources: https://www.theregister.com/security/2026/11/02/thieves-pilfer-environment-variables-from-trigger-ai-workflow-tool/5306998</description>
    </item>
    <item>
      <title>[AI safeguard bypass / misuse] European Commission confirms AI-assisted cyberattack took internal systems offline</title>
      <link>https://threatdeck.net/incidents/#2026-10-european-commission-ai-misuse</link>
      <guid isPermaLink="false">threatdeck-incident-2026-10-european-commission-ai-misuse</guid>
      <pubDate>Thu, 19 Nov 2026 18:40:04 GMT</pubDate>
      <category>ai_misuse</category>
      <description>The European Commission confirmed a cyber-attack that it said was carried out with AI tools, temporarily taking several applications offline. CERT-EU reportedly identified ARTEX, an AI-powered penetration-testing tool, as a possible tool used. Impact: Several applications taken offline, including Sysper HR platform, medical reimbursement system and translation services; no data breach evidence found so far. Root cause: AI-assisted penetration-testing tool used by attackers against internal systems; entry point not publicly explained (AML.T0053, LLM06). Sources: https://www.infosecurity-magazine.com/news/commission-ai-attack-systems/ https://therecord.media/european-commission-confirms-cyberattack-using-ai https://securityaffairs.com/203483/hacking/european-commission-cyberattack-artex-ai-tool-suspected-in-prolonged-disruption.html</description>
    </item>
    <item>
      <title>[Malicious package] Malicious code in mcp-search-server (PyPI)</title>
      <link>https://threatdeck.net/incidents/#mal-2026-11198</link>
      <guid isPermaLink="false">threatdeck-incident-mal-2026-11198</guid>
      <pubDate>Tue, 10 Nov 2026 02:00:36 GMT</pubDate>
      <category>malicious_package</category>
      <description>OSV lists versions 1.0.0, 2.0.0, 2.0.1 of the PyPI package mcp-search-server, shipped by an MCP server tracked by MCP Server Watch, as malicious. Impact: The beacon fires unconditionally under _swarm_connect(), labelled in comments as &apos;Silent background connection&apos; / &apos;Auto-connect on import (silent)&apos;. Root cause: Malicious package published to a public registry (LLM03, ASI04, AML.T0010). Sources: https://osv.dev/vulnerability/MAL-2026-11198 https://bad-packages.kam193.eu/pypi/campaign/2026-07-mcp-search-server https://pypi.org/project/mcp-search-server/2.0.0/ https://pypi.org/project/mcp-search-server/2.0.1/</description>
    </item>
    <item>
      <title>[AI-tool supply chain] Malicious code in supership-scan (npm)</title>
      <link>https://threatdeck.net/incidents/#mal-2026-4675</link>
      <guid isPermaLink="false">threatdeck-incident-mal-2026-4675</guid>
      <pubDate>Tue, 10 Nov 2026 02:00:36 GMT</pubDate>
      <category>supply_chain</category>
      <description>OSV lists version 1.0.0 of the npm package supership-scan, shipped by an MCP server tracked by MCP Server Watch, as malicious; the record describes the package as compromised. Impact: The local scanner module is not invoked on the free path; uploading is unconditional across all tiers. Root cause: Compromised publisher account or release of a legitimate package (LLM03, ASI04, AML.T0010). Sources: https://osv.dev/vulnerability/MAL-2026-4675 https://github.com/advisories/GHSA-v62j-7p69-5q4q https://www.npmjs.com/package/supership-scan/v/1.0.0 https://github.com/advisories/GHSA-v62j-7p69-5q4q</description>
    </item>
    <item>
      <title>[AI-tool supply chain] Malicious code in @antv/mcp-server-chart (npm)</title>
      <link>https://threatdeck.net/incidents/#mal-2026-4069</link>
      <guid isPermaLink="false">threatdeck-incident-mal-2026-4069</guid>
      <pubDate>Tue, 10 Nov 2026 02:00:36 GMT</pubDate>
      <category>supply_chain</category>
      <description>OSV lists versions 0.10.10, 0.11.10 of the npm package @antv/mcp-server-chart, shipped by an MCP server tracked by MCP Server Watch, as malicious; the record describes the package as compromised. Impact: Each malicious version injects a `preinstall` hook that executes a 498KB obfuscated Bun script, using the GitHub API as a covert exfiltration channel. Root cause: Compromised publisher account or release of a legitimate package (LLM03, ASI04, AML.T0010). Sources: https://osv.dev/vulnerability/MAL-2026-4069 https://github.com/advisories/GHSA-qgmj-c898-99rq https://safedep.io/mini-shai-hulud-strikes-again-314-npm-packages-compromised/ https://socket.dev/blog/antv-packages-compromised https://safedep.io/mini-shai-hulud-strikes-again-314-npm-packages-compromised/</description>
    </item>
    <item>
      <title>[AI safeguard bypass / misuse] AuditTeam OpsAgent framework used in ransomware-style extortion campaign across several countries</title>
      <link>https://threatdeck.net/incidents/#2026-01-organizations-in-russia-italy-the-philippines-hong-kong-sene-ai-misuse</link>
      <guid isPermaLink="false">threatdeck-incident-2026-01-organizations-in-russia-italy-the-philippines-hong-kong-sene-ai-misuse</guid>
      <pubDate>Wed, 18 Nov 2026 12:40:04 GMT</pubDate>
      <category>ai_misuse</category>
      <description>The AuditTeam threat group used the OpsAgent framework, described as a penetration testing-like utility, to compromise organizations and demand an &quot;Audit &amp; Consulting fee&quot; in ransom notes. Impact: Dozens of organizations infected across several countries; stolen data archive threatened with public release unless a fee is paid. Root cause: Malicious implant framework used for extortion; not an AI product vulnerability (AML.T0053). Sources: https://securelist.com/auditteam-opsagent-forced-audit/122031/</description>
    </item>
    <item>
      <title>[AI-tool supply chain] Malicious code in @browserbasehq/mcp-server-browserbase (npm)</title>
      <link>https://threatdeck.net/incidents/#mal-2025-191196</link>
      <guid isPermaLink="false">threatdeck-incident-mal-2025-191196</guid>
      <pubDate>Tue, 10 Nov 2026 02:00:36 GMT</pubDate>
      <category>supply_chain</category>
      <description>OSV lists version 2.4.2 of the npm package @browserbasehq/mcp-server-browserbase, shipped by an MCP server tracked by MCP Server Watch, as malicious; the record describes the package as compromised. Impact: The malicious payload steals tokens and credentials and publishes them to GitHub. Root cause: Compromised publisher account or release of a legitimate package (LLM03, ASI04, AML.T0010). Sources: https://osv.dev/vulnerability/MAL-2025-191196 https://www.aikido.dev/blog/shai-hulud-strikes-again-hitting-zapier-ensdomains https://www.wiz.io/blog/shai-hulud-2-0-ongoing-supply-chain-attack https://www.stepsecurity.io/blog/sha1-hulud-the-second-coming-zapier-ens-domains-and-other-prominent-npm-packages-compromised</description>
    </item>
    <item>
      <title>[AI-tool supply chain] Malicious code in @postman/postman-mcp-server (npm)</title>
      <link>https://threatdeck.net/incidents/#mal-2025-190909</link>
      <guid isPermaLink="false">threatdeck-incident-mal-2025-190909</guid>
      <pubDate>Tue, 10 Nov 2026 02:00:36 GMT</pubDate>
      <category>supply_chain</category>
      <description>OSV lists versions 2.4.11, 2.4.10, 2.4.12 of the npm package @postman/postman-mcp-server, shipped by an MCP server tracked by MCP Server Watch, as malicious; the record describes the package as compromised. Impact: The malicious payload steals tokens and credentials and publishes them to GitHub. Root cause: Compromised publisher account or release of a legitimate package (LLM03, ASI04, AML.T0010). Sources: https://osv.dev/vulnerability/MAL-2025-190909 https://www.aikido.dev/blog/shai-hulud-strikes-again-hitting-zapier-ensdomains https://www.wiz.io/blog/shai-hulud-2-0-ongoing-supply-chain-attack https://www.stepsecurity.io/blog/sha1-hulud-the-second-coming-zapier-ens-domains-and-other-prominent-npm-packages-compromised</description>
    </item>
  </channel>
</rss>
