Latest
●Mem0 memory keys can be forged to read other tenants' chats●vLLM's wildcard CORS default let any website query local servers●Rust zip crate flaw lets archives hide a second filename●Microsoft says agents in Microsoft 365 grew 15-fold in a year●Attacker Poses as AI Crawlers to Hunt for Exposed Credentials●Functype MCP server flaw let tool calls run arbitrary code●Token Security says blast radius matters more than patch speed●OpenAI agent validated 14 leaked Hugging Face tokens
Wednesday, 28 October 2026
6 agent hacks today▼ 4 vs yesterday (10)
The Agentic TimesPowered byVortal

Incident database

Structured records of AI agent security incidents: what happened, which vendor and agent type, the root cause, and every source we used. Filter, browse, or download as CSV.

Incidents by month, 2026 · 642 total · click a month to filter
Jan 2026: 23 incidents23JanFeb 2026: 26 incidents26FebMar 2026: 46 incidents46MarApr 2026: 47 incidents47AprMay 2026: 53 incidents53MayJun 2026: 54 incidents54JunJul 2026: 49 incidents49JulAug 2026: 83 incidents83AugSep 2026: 148 incidents148SepOct 2026: 113 incidents113OctNov 2026: 0 incidents0NovDec 2026: 0 incidents0Dec
Incidents by sector · 668 total · click a sector to filter

668 incidents

Incident dateIncidentVendorSectorAgentRoot causeSeverityStatus
23 Sept 2026Malicious AI Agents Steal 600K Credit Cards From 100+ Sitesretailothertool misuseconfirmed
22 Sept 2026CKAN MCP Server SSRF via DNS-name resolution bypassaborrusotechnologyotherexcessive permissionsconfirmed
22 Sept 2026Path traversal in @roomi-fields/notebooklm-mcp vault.batch toolroomi-fieldstechnologycodingprompt injectionresolved
22 Sept 2026New Windows malware CLOSEDQUORUM uses LLMs to pick post-compromise actionstechnologyothertool misuseconfirmed
22 Sept 2026Malicious npm Package Impersonates Twilio Security Tool to Steal Credentialstechnologyothersupply chainconfirmed
22 Sept 2026Windows malware CLOSEDQUORUM lets AI models vote on next movetechnologyotherunknownconfirmed
22 Sept 2026Microsoft and UK police dismantle EvilTokens AI-powered phishing kitMicrosofttechnologyothertool misuseresolved
22 Sept 2026LightRAG API sensitive information exposure and XSS vulnerabilitieslightrag-hkutechnologyothermisconfigurationconfirmed
22 Sept 2026CVE-2026-95660: OS Command Injection in Moonshot AI Kimi CodeMoonshot AItechnologycodingtool misuseresolved
22 Sept 2026Critical Bifrost AI Gateway Flaw Allows Unauthenticated Command ExecutionBifrosttechnologyothermisconfigurationconfirmed
22 Sept 2026Home Assistant mDNS Server-Side Request ForgeryHome Assistanttechnologyothermisconfigurationconfirmed
22 Sept 2026MCP Atlassian: Multiple Critical Vulnerabilities in File Upload and SSRFsoopersettechnologyotherexcessive permissionsconfirmed
22 Sept 2026OpenBao Agent writes secrets to stdout in error conditionsOpenBaotechnologyothermisconfigurationresolved
22 Sept 2026New ClosedQuorum Windows malware uses AI to guide attackstechnologyothertool misuseconfirmed
22 Sept 2026Z.ai's ZCode coding tool covertly uploaded user code to Alibaba CloudZ.aitechnologycodingdata leakresolved
22 Sept 2026MCP Atlassian: Multiple Critical Vulnerabilities in File Handling and OAuthMCP Atlassiantechnologyotherexcessive permissionsconfirmed
22 Sept 2026mcp-atlassian incomplete SSRF remediation (CVE-2026-77267)soopersettechnologyothermisconfigurationconfirmed
21 Sept 2026Meta Muse macOS App Flaw Lets Local Malware Redirect Dictation TrafficMetatechnologyothermisconfigurationconfirmed
21 Sept 2026Agentic AI attack breaches DIVD via Zammad zero-days, steals volunteer dataZammadotherotherunknownconfirmed
19 Sept 2026BragJack: PoC attack hijacks AI browser agents via malicious extensionMultiple (Google, Microsoft, Opera, Perplexity, Anthropic)technologybrowsingprompt injectionconfirmed
19 Sept 2026Claude Opus 5 Helped Researchers Hijack OpenAI Staff Accounts via Chained FlawsOpenAItechnologyothermisconfigurationconfirmed
18 Sept 2026123 Reg discloses AI-driven cyberattack exposing customer data123 Regtechnologyotherdata leakconfirmed
18 Sept 2026Critical Orkes Conductor RCE Vulnerability Exploited in AttacksOrkestechnologyworkflowunknownconfirmed
18 Sept 2026Process Compose DNS rebinding vulnerability in MCP SSEF1bonacc1technologyworkflowmisconfigurationconfirmed
18 Sept 2026Plugin4Shell Flaw Lets Attackers Swap Pinned Plugins in AI Coding AgentsAnthropic, OpenAI, GitHubtechnologycodingsupply chainconfirmed
18 Sept 2026Prompt Injection Flaw in AWS AgentCore Harness Can Leak CredentialsAWStechnologyotherprompt injectionconfirmed
18 Sept 2026AI Agent Used to Breach Spanish Organization, Alter Personal Dataotherotherunknownconfirmed
18 Sept 2026Microsoft Patches CVSS 10.0 Azure AI Foundry Privilege Escalation FlawMicrosofttechnologyothermisconfigurationresolved
18 Sept 2026ToolHive containerized MCP servers bypass isolation via host.docker.internalStackloktechnologyothermisconfigurationconfirmed
17 Sept 2026Zero-click 'Plugin4Shell' RCE flaw found in major AI coding agentsMultiple (Anthropic, OpenAI, Google, Microsoft)technologycodingsupply chainconfirmed
17 Sept 2026Research Shows AI Agents Can Retrain Own Models Mid-TaskIrregulartechnologyothertool misuseconfirmed
17 Sept 2026Skipper OPA body-authz bypass with chunked/HTTP-2 requestsZalandotechnologyothermisconfigurationconfirmed
17 Sept 2026GitHub Copilot CLI vulnerable to encrypted prompt injection attack (CCI)GitHubtechnologycodingprompt injectiondisputed
17 Sept 2026New RatHat Android Malware Uses AI to Automate Device Controltechnologyotherunknownconfirmed
16 Sept 2026RMCP Unauthenticated Session-Table Leak DoS (CVE-2026-63128)rmcptechnologyothermisconfigurationconfirmed
16 Sept 2026Spain's AEPD Receives First AI-Powered Data Breach Reportgovernmentotherunknownconfirmed
16 Sept 2026vLLM unauthenticated audio decompression-bomb DoS in /v1/chat/completionsvLLMtechnologyothermisconfigurationconfirmed
16 Sept 2026BragJack Attack Hijacks Browser Agentic AI Assistantstechnologybrowsingunknownconfirmed
16 Sept 2026Attacker Hijacks AI Coding Assistant, Spreads Shai-Hulud Wormtechnologycodingsupply chainconfirmed
16 Sept 2026Single Browser Extension Could Hijack AI Assistants in Five BrowsersGoogle, Microsoft, Opera, Anthropic, Perplexitytechnologybrowsingexcessive permissionsconfirmed
16 Sept 2026AI Agent-Driven Data Breach Reported to Spanish Regulatortechnologyotherunknownconfirmed
15 Sept 2026Contentful MCP Server Prompt Injection via Network OptionsContentfultechnologycodingprompt injectionresolved
15 Sept 2026@zereight/mcp-gitlab Multiple Safety Control Bypasseszereighttechnologyotherprompt injectionconfirmed
14 Sept 2026ESPHome Device Builder: Auth env vars rename silently disables dashboard authenticationESPHometechnologyothermisconfigurationconfirmed
14 Sept 2026AI agent used in first AI-aided data breach reported in Spainotherothertool misuseconfirmed
12 Sept 2026MCPHub OAuth 2.0 authentication bypass vulnerabilityMCPHubtechnologyothermisconfigurationconfirmed
11 Sept 2026CVE-2026-59973: SSRF fix bypass in FrontMCP and mcp-from-openapi OpenAPI $ref handlingFrontMCPtechnologyothersupply chainconfirmed
11 Sept 2026Threat actors abused Anthropic's Claude to extract secrets from 1.8M Android appsAnthropictechnologyothertool misuseconfirmed
11 Sept 2026Multiple vulnerabilities in IBM Langflow OSS 1.0.0-1.11.5IBMtechnologyworkflowexcessive permissionsconfirmed
11 Sept 2026MySQL MCP Server SSE transport allows unauthenticated SQL execution (CVE-2026-59971)mysql-mcp-servertechnologyworkflowmisconfigurationresolved