[](https://www.paloaltonetworks.com/?ts=markdown)

* Sign In
  
  * Customer
  * Partner
  * Employee
  * [Login to download](https://www.paloaltonetworks.com/login?ts=markdown)
  * [Join us to become a member](https://www.paloaltonetworks.com/login?screenToRender=traditionalRegistration&ts=markdown)

* EN
  
  * [USA (ENGLISH)](https://www.paloaltonetworks.com)
  * [AUSTRALIA (ENGLISH)](https://www.paloaltonetworks.com.au)
  * [BRAZIL (PORTUGUÉS)](https://www.paloaltonetworks.com.br)
  * [CANADA (ENGLISH)](https://www.paloaltonetworks.ca)
  * [CHINA (简体中文)](https://www.paloaltonetworks.cn)
  * [FRANCE (FRANÇAIS)](https://www.paloaltonetworks.fr)
  * [GERMANY (DEUTSCH)](https://www.paloaltonetworks.de)
  * [INDIA (ENGLISH)](https://www.paloaltonetworks.in)
  * [ITALY (ITALIANO)](https://www.paloaltonetworks.it)
  * [JAPAN (日本語)](https://www.paloaltonetworks.jp)
  * [KOREA (한국어)](https://www.paloaltonetworks.co.kr)
  * [LATIN AMERICA (ESPAÑOL)](https://www.paloaltonetworks.lat)
  * [MEXICO (ESPAÑOL)](https://www.paloaltonetworks.com.mx)
  * [SINGAPORE (ENGLISH)](https://www.paloaltonetworks.sg)
  * [SPAIN (ESPAÑOL)](https://www.paloaltonetworks.es)
  * [TAIWAN (繁體中文)](https://www.paloaltonetworks.tw)
  * [UK (ENGLISH)](https://www.paloaltonetworks.co.uk)

* ![magnifying glass search icon to open search field](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/search-black.svg)

* [Contact Us](https://www.paloaltonetworks.com/company/contact-sales?ts=markdown)

* [What's New](https://www.paloaltonetworks.com/resources?ts=markdown)

* [Get Support](https://support.paloaltonetworks.com/SupportAccount/MyAccount)

* [Under Attack?](https://start.paloaltonetworks.com/contact-unit42.html)
  ![x close icon to close mobile navigation](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/x-black.svg) [![Palo Alto Networks logo](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/pan-logo-dark.svg)](https://www.paloaltonetworks.com/?ts=markdown) ![magnifying glass search icon to open search field](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/search-black.svg)

* [](https://www.paloaltonetworks.com/?ts=markdown)

* Products  
  ![black arrow pointing left to go back to main navigation](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/arrow-right-black.svg) Products  
  [AI-Powered Network Security Platform](https://www.paloaltonetworks.com/network-security?ts=markdown)
  
  * [AI Security](https://www.paloaltonetworks.com/precision-ai-security/secure-ai-by-design?ts=markdown)
  
  * [Prisma AIRS](https://www.paloaltonetworks.com/prisma/prisma-ai-runtime-security?ts=markdown)
  
  * [AI Access Security](https://www.paloaltonetworks.com/sase/ai-access-security?ts=markdown)
  
  * [Cloud Delivered Security Services](https://www.paloaltonetworks.com/network-security/security-subscriptions?ts=markdown)
  
  * [Advanced Threat Prevention](https://www.paloaltonetworks.com/network-security/advanced-threat-prevention?ts=markdown)
  
  * [Advanced URL Filtering](https://www.paloaltonetworks.com/network-security/advanced-url-filtering?ts=markdown)
  
  * [Advanced WildFire](https://www.paloaltonetworks.com/network-security/advanced-wildfire?ts=markdown)
  
  * [Advanced DNS Security](https://www.paloaltonetworks.com/network-security/advanced-dns-security?ts=markdown)
  
  * [Enterprise Data Loss Prevention](https://www.paloaltonetworks.com/sase/enterprise-data-loss-prevention?ts=markdown)
  
  * [Enterprise Device Security](https://www.paloaltonetworks.com/network-security/enterprise-device-security?ts=markdown)
  
  * [Medical Device Security](https://www.paloaltonetworks.com/network-security/medical-device-security?ts=markdown)
  
  * [OT Device Security](https://www.paloaltonetworks.com/network-security/ot-device-security?ts=markdown)
  
  * [SaaS Security](https://www.paloaltonetworks.com/sase/saas-security?ts=markdown)
  
  * [Next-Generation Firewalls](https://www.paloaltonetworks.com/network-security/next-generation-firewall?ts=markdown)
  
  * [Hardware Firewalls](https://www.paloaltonetworks.com/network-security/hardware-firewall-innovations?ts=markdown)
  
  * [Software Firewalls](https://www.paloaltonetworks.com/network-security/software-firewalls?ts=markdown)
  
  * [Strata Cloud Manager](https://www.paloaltonetworks.com/network-security/strata-cloud-manager?ts=markdown)
  
  * [SD-WAN for NGFW](https://www.paloaltonetworks.com/network-security/sd-wan-subscription?ts=markdown)
  
  * [PAN-OS](https://www.paloaltonetworks.com/network-security/pan-os?ts=markdown)
  
  * [Panorama](https://www.paloaltonetworks.com/network-security/panorama?ts=markdown)
  
  * [Secure Access Service Edge](https://www.paloaltonetworks.com/sase?ts=markdown)
  
  * [Prisma SASE](https://www.paloaltonetworks.com/sase?ts=markdown)
  
  * [Application Acceleration](https://www.paloaltonetworks.com/sase/app-acceleration?ts=markdown)
  
  * [Autonomous Digital Experience Management](https://www.paloaltonetworks.com/sase/adem?ts=markdown)
  
  * [Enterprise DLP](https://www.paloaltonetworks.com/sase/enterprise-data-loss-prevention?ts=markdown)
  
  * [Prisma Access](https://www.paloaltonetworks.com/sase/access?ts=markdown)
  
  * [Prisma Browser](https://www.paloaltonetworks.com/sase/prisma-browser?ts=markdown)
  
  * [Prisma SD-WAN](https://www.paloaltonetworks.com/sase/sd-wan?ts=markdown)
  
  * [Remote Browser Isolation](https://www.paloaltonetworks.com/sase/remote-browser-isolation?ts=markdown)
  
  * [SaaS Security](https://www.paloaltonetworks.com/sase/saas-security?ts=markdown)  
    [AI-Driven Security Operations Platform](https://www.paloaltonetworks.com/cortex?ts=markdown)
  
  * [Cloud Security](https://www.paloaltonetworks.com/cortex/cloud?ts=markdown)
  
  * [Cortex Cloud](https://www.paloaltonetworks.com/cortex/cloud?ts=markdown)
  
  * [Application Security](https://www.paloaltonetworks.com/cortex/cloud/application-security?ts=markdown)
  
  * [Cloud Posture Security](https://www.paloaltonetworks.com/cortex/cloud/cloud-posture-security?ts=markdown)
  
  * [Cloud Runtime Security](https://www.paloaltonetworks.com/cortex/cloud/runtime-security?ts=markdown)
  
  * [Prisma Cloud](https://www.paloaltonetworks.com/prisma/cloud?ts=markdown)
  
  * [AI-Driven SOC](https://www.paloaltonetworks.com/cortex?ts=markdown)
  
  * [Cortex XSIAM](https://www.paloaltonetworks.com/cortex/cortex-xsiam?ts=markdown)
  
  * [Cortex XDR](https://www.paloaltonetworks.com/cortex/cortex-xdr?ts=markdown)
  
  * [Cortex AgentiX](https://www.paloaltonetworks.com/cortex/agentix?ts=markdown)
  
  * [Cortex XSOAR](https://www.paloaltonetworks.com/cortex/cortex-xsoar?ts=markdown)
  
  * [Cortex Exposure Management](https://www.paloaltonetworks.com/cortex/exposure-management?ts=markdown)
  
  * [Cortex Xpanse](https://www.paloaltonetworks.com/cortex/cortex-xpanse?ts=markdown)
  
  * [Cortex Advanced Email Security](https://www.paloaltonetworks.com/cortex/advanced-email-security?ts=markdown)
  
  * [Unit 42 Managed Detection \& Response](https://www.paloaltonetworks.com/cortex/managed-detection-and-response?ts=markdown)
  
  * [Unit 42 Managed XSIAM](https://www.paloaltonetworks.com/cortex/managed-xsiam?ts=markdown)

* Solutions  
  ![black arrow pointing left to go back to main navigation](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/arrow-right-black.svg) Solutions  
  Secure AI by Design
  
  * [Secure AI Ecosystem](https://www.paloaltonetworks.com/prisma/prisma-ai-runtime-security?ts=markdown)
  * [Secure GenAI Usage](https://www.paloaltonetworks.com/sase/ai-access-security?ts=markdown)  
    Network Security
  * [Cloud Network Security](https://www.paloaltonetworks.com/network-security/software-firewalls?ts=markdown)
  * [Data Center Security](https://www.paloaltonetworks.com/network-security/data-center?ts=markdown)
  * [DNS Security](https://www.paloaltonetworks.com/network-security/advanced-dns-security?ts=markdown)
  * [Intrusion Detection and Prevention](https://www.paloaltonetworks.com/network-security/advanced-threat-prevention?ts=markdown)
  * [Device Security](https://www.paloaltonetworks.com/network-security/device-security?ts=markdown)
  * [OT Security](https://www.paloaltonetworks.com/network-security/ot-device-security?ts=markdown)
  * [5G Security](https://www.paloaltonetworks.com/network-security/5g-security?ts=markdown)
  * [Secure All Apps, Users and Locations](https://www.paloaltonetworks.com/sase/secure-users-data-apps-devices?ts=markdown)
  * [Secure Branch Transformation](https://www.paloaltonetworks.com/sase/secure-branch-transformation?ts=markdown)
  * [Secure Work on Any Device](https://www.paloaltonetworks.com/sase/secure-work-on-any-device?ts=markdown)
  * [VPN Replacement](https://www.paloaltonetworks.com/sase/vpn-replacement-for-secure-remote-access?ts=markdown)
  * [Web \& Phishing Security](https://www.paloaltonetworks.com/network-security/advanced-url-filtering?ts=markdown)  
    Cloud Security
  * [Application Security Posture Management (ASPM)](https://www.paloaltonetworks.com/cortex/cloud/application-security-posture-management?ts=markdown)
  * [Software Supply Chain Security](https://www.paloaltonetworks.com/cortex/cloud/software-supply-chain-security?ts=markdown)
  * [Code Security](https://www.paloaltonetworks.com/cortex/cloud/code-security?ts=markdown)
  * [Cloud Security Posture Management (CSPM)](https://www.paloaltonetworks.com/cortex/cloud/cloud-security-posture-management?ts=markdown)
  * [Cloud Infrastructure Entitlement Management (CIEM)](https://www.paloaltonetworks.com/cortex/cloud/cloud-infrastructure-entitlement-management?ts=markdown)
  * [Data Security Posture Management (DSPM)](https://www.paloaltonetworks.com/cortex/cloud/data-security-posture-management?ts=markdown)
  * [AI Security Posture Management (AI-SPM)](https://www.paloaltonetworks.com/cortex/cloud/ai-security-posture-management?ts=markdown)
  * [Cloud Detection \& Response](https://www.paloaltonetworks.com/cortex/cloud-detection-and-response?ts=markdown)
  * [Cloud Workload Protection (CWP)](https://www.paloaltonetworks.com/cortex/cloud/cloud-workload-protection?ts=markdown)
  * [Web Application \& API Security (WAAS)](https://www.paloaltonetworks.com/cortex/cloud/web-app-api-security?ts=markdown)  
    Security Operations
  * [Cloud Detection \& Response](https://www.paloaltonetworks.com/cortex/cloud-detection-and-response?ts=markdown)
  * [Security Information and Event Management](https://www.paloaltonetworks.com/cortex/modernize-siem?ts=markdown)
  * [Network Security Automation](https://www.paloaltonetworks.com/cortex/network-security-automation?ts=markdown)
  * [Incident Case Management](https://www.paloaltonetworks.com/cortex/incident-case-management?ts=markdown)
  * [SOC Automation](https://www.paloaltonetworks.com/cortex/security-operations-automation?ts=markdown)
  * [Threat Intel Management](https://www.paloaltonetworks.com/cortex/threat-intel-management?ts=markdown)
  * [Managed Detection \& Response](https://www.paloaltonetworks.com/cortex/managed-detection-and-response?ts=markdown)
  * [Attack Surface Management](https://www.paloaltonetworks.com/cortex/cortex-xpanse/attack-surface-management?ts=markdown)
  * [Compliance Management](https://www.paloaltonetworks.com/cortex/cortex-xpanse/compliance-management?ts=markdown)
  * [Internet Operations Management](https://www.paloaltonetworks.com/cortex/cortex-xpanse/internet-operations-management?ts=markdown)
  * [Extended Data Lake (XDL)](https://www.paloaltonetworks.com/cortex/cortex-xdl?ts=markdown)
  * [Agentic Assistant](https://www.paloaltonetworks.com/cortex/cortex-agentic-assistant?ts=markdown)  
    Endpoint Security
  * [Endpoint Protection](https://www.paloaltonetworks.com/cortex/endpoint-protection?ts=markdown)
  * [Extended Detection \& Response](https://www.paloaltonetworks.com/cortex/detection-and-response?ts=markdown)
  * [Ransomware Protection](https://www.paloaltonetworks.com/cortex/ransomware-protection?ts=markdown)
  * [Digital Forensics](https://www.paloaltonetworks.com/cortex/digital-forensics?ts=markdown)  
    [Industries](https://www.paloaltonetworks.com/industry?ts=markdown)
  * [Public Sector](https://www.paloaltonetworks.com/industry/public-sector?ts=markdown)
  * [Financial Services](https://www.paloaltonetworks.com/industry/financial-services?ts=markdown)
  * [Manufacturing](https://www.paloaltonetworks.com/industry/manufacturing?ts=markdown)
  * [Healthcare](https://www.paloaltonetworks.com/industry/healthcare?ts=markdown)
  * [Small \& Medium Business Solutions](https://www.paloaltonetworks.com/industry/small-medium-business-portfolio?ts=markdown)

* Services  
  ![black arrow pointing left to go back to main navigation](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/arrow-right-black.svg) Services  
  [Threat Intel and Incident Response Services](https://www.paloaltonetworks.com/unit42?ts=markdown)
  
  * [Assess](https://www.paloaltonetworks.com/unit42/assess?ts=markdown)
  
  * [AI Security Assessment](https://www.paloaltonetworks.com/unit42/assess/ai-security-assessment?ts=markdown)
  
  * [Attack Surface Assessment](https://www.paloaltonetworks.com/unit42/assess/attack-surface-assessment?ts=markdown)
  
  * [Breach Readiness Review](https://www.paloaltonetworks.com/unit42/assess/breach-readiness-review?ts=markdown)
  
  * [BEC Readiness Assessment](https://www.paloaltonetworks.com/bec-readiness-assessment?ts=markdown)
  
  * [Cloud Security Assessment](https://www.paloaltonetworks.com/unit42/assess/cloud-security-assessment?ts=markdown)
  
  * [Compromise Assessment](https://www.paloaltonetworks.com/unit42/assess/compromise-assessment?ts=markdown)
  
  * [Cyber Risk Assessment](https://www.paloaltonetworks.com/unit42/assess/cyber-risk-assessment?ts=markdown)
  
  * [M\&A Cyber Due Diligence](https://www.paloaltonetworks.com/unit42/assess/mergers-acquisitions-cyber-due-diligence?ts=markdown)
  
  * [Penetration Testing](https://www.paloaltonetworks.com/unit42/assess/penetration-testing?ts=markdown)
  
  * [Purple Team Exercises](https://www.paloaltonetworks.com/unit42/assess/purple-teaming?ts=markdown)
  
  * [Ransomware Readiness Assessment](https://www.paloaltonetworks.com/unit42/assess/ransomware-readiness-assessment?ts=markdown)
  
  * [SOC Assessment](https://www.paloaltonetworks.com/unit42/assess/soc-assessment?ts=markdown)
  
  * [Supply Chain Risk Assessment](https://www.paloaltonetworks.com/unit42/assess/supply-chain-risk-assessment?ts=markdown)
  
  * [Tabletop Exercises](https://www.paloaltonetworks.com/unit42/assess/tabletop-exercise?ts=markdown)
  
  * [Unit 42 Retainer](https://www.paloaltonetworks.com/unit42/retainer?ts=markdown)
  
  * [Respond](https://www.paloaltonetworks.com/unit42/respond?ts=markdown)
  
  * [Cloud Incident Response](https://www.paloaltonetworks.com/unit42/respond/cloud-incident-response?ts=markdown)
  
  * [Digital Forensics](https://www.paloaltonetworks.com/unit42/respond/digital-forensics?ts=markdown)
  
  * [Incident Response](https://www.paloaltonetworks.com/unit42/respond/incident-response?ts=markdown)
  
  * [Managed Detection and Response](https://www.paloaltonetworks.com/unit42/respond/managed-detection-response?ts=markdown)
  
  * [Managed Threat Hunting](https://www.paloaltonetworks.com/unit42/respond/managed-threat-hunting?ts=markdown)
  
  * [Managed XSIAM](https://www.paloaltonetworks.com/cortex/managed-xsiam?ts=markdown)
  
  * [Unit 42 Retainer](https://www.paloaltonetworks.com/unit42/retainer?ts=markdown)
  
  * [Transform](https://www.paloaltonetworks.com/unit42/transform?ts=markdown)
  
  * [IR Plan Development and Review](https://www.paloaltonetworks.com/unit42/transform/incident-response-plan-development-review?ts=markdown)
  
  * [Security Program Design](https://www.paloaltonetworks.com/unit42/transform/security-program-design?ts=markdown)
  
  * [Virtual CISO](https://www.paloaltonetworks.com/unit42/transform/vciso?ts=markdown)
  
  * [Zero Trust Advisory](https://www.paloaltonetworks.com/unit42/transform/zero-trust-advisory?ts=markdown)  
    [Global Customer Services](https://www.paloaltonetworks.com/services?ts=markdown)
  
  * [Education \& Training](https://www.paloaltonetworks.com/services/education?ts=markdown)
  
  * [Professional Services](https://www.paloaltonetworks.com/services/consulting?ts=markdown)
  
  * [Success Tools](https://www.paloaltonetworks.com/services/customer-success-tools?ts=markdown)
  
  * [Support Services](https://www.paloaltonetworks.com/services/solution-assurance?ts=markdown)
  
  * [Customer Success](https://www.paloaltonetworks.com/services/customer-success?ts=markdown)  
    [![](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/logo-unit-42.svg)  
    UNIT 42 RETAINER
    Custom-built to fit your organization's needs, you can choose to allocate your retainer hours to any of our offerings, including proactive cyber risk management services. Learn how you can put the world-class Unit 42 Incident Response team on speed dial.
    Learn more](https://www.paloaltonetworks.com/unit42/retainer?ts=markdown)

* Partners  
  ![black arrow pointing left to go back to main navigation](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/arrow-right-black.svg) Partners  
  NextWave Partners
  
  * [NextWave Partner Community](https://www.paloaltonetworks.com/partners?ts=markdown)
  * [Cloud Service Providers](https://www.paloaltonetworks.com/partners/nextwave-for-csp?ts=markdown)
  * [Global Systems Integrators](https://www.paloaltonetworks.com/partners/nextwave-for-gsi?ts=markdown)
  * [Technology Partners](https://www.paloaltonetworks.com/partners/technology-partners?ts=markdown)
  * [Service Providers](https://www.paloaltonetworks.com/partners/service-providers?ts=markdown)
  * [Solution Providers](https://www.paloaltonetworks.com/partners/nextwave-solution-providers?ts=markdown)
  * [Managed Security Service Providers](https://www.paloaltonetworks.com/partners/managed-security-service-providers?ts=markdown)
  * [XMDR Partners](https://www.paloaltonetworks.com/partners/managed-security-service-providers/xmdr?ts=markdown)  
    Take Action
  * [Portal Login](https://www.paloaltonetworks.com/partners/nextwave-partner-portal?ts=markdown)
  * [Managed Services Program](https://www.paloaltonetworks.com/partners/managed-security-services-provider-program?ts=markdown)
  * [Become a Partner](https://paloaltonetworks.my.site.com/NextWavePartnerProgram/s/partnerregistration?type=becomepartner)
  * [Request Access](https://paloaltonetworks.my.site.com/NextWavePartnerProgram/s/partnerregistration?type=requestaccess)
  * [Find a Partner](https://paloaltonetworks.my.site.com/NextWavePartnerProgram/s/partnerlocator)  
    [CYBERFORCE
    CYBERFORCE represents the top 1% of partner engineers trusted for their security expertise.
    Learn more](https://www.paloaltonetworks.com/cyberforce?ts=markdown)

* Company  
  ![black arrow pointing left to go back to main navigation](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/arrow-right-black.svg) Company  
  Palo Alto Networks
  
  * [About Us](https://www.paloaltonetworks.com/about-us?ts=markdown)
  * [Management Team](https://www.paloaltonetworks.com/about-us/management?ts=markdown)
  * [Investor Relations](https://investors.paloaltonetworks.com)
  * [Locations](https://www.paloaltonetworks.com/about-us/locations?ts=markdown)
  * [Ethics \& Compliance](https://www.paloaltonetworks.com/company/ethics-and-compliance?ts=markdown)
  * [Corporate Responsibility](https://www.paloaltonetworks.com/about-us/corporate-responsibility?ts=markdown)
  * [Military \& Veterans](https://jobs.paloaltonetworks.com/military)  
    [Why Palo Alto Networks?](https://www.paloaltonetworks.com/why-paloaltonetworks?ts=markdown)
  * [Precision AI Security](https://www.paloaltonetworks.com/precision-ai-security?ts=markdown)
  * [Our Platform Approach](https://www.paloaltonetworks.com/why-paloaltonetworks/platformization?ts=markdown)
  * [Accelerate Your Cybersecurity Transformation](https://www.paloaltonetworks.com/why-paloaltonetworks/nam-cxo-portfolio?ts=markdown)
  * [Awards \& Recognition](https://www.paloaltonetworks.com/about-us/awards?ts=markdown)
  * [Customer Stories](https://www.paloaltonetworks.com/customers?ts=markdown)
  * [Global Certifications](https://www.paloaltonetworks.com/legal-notices/trust-center/compliance?ts=markdown)
  * [Trust 360 Program](https://www.paloaltonetworks.com/resources/whitepapers/trust-360?ts=markdown)  
    Careers
  * [Overview](https://jobs.paloaltonetworks.com/)
  * [Culture \& Benefits](https://jobs.paloaltonetworks.com/en/culture/)  
    [A Newsweek Most Loved Workplace
    "Businesses that do right by their employees"
    Read more](https://www.paloaltonetworks.com/company/press/2021/palo-alto-networks-secures-top-ranking-on-newsweek-s-most-loved-workplaces-list-for-2021?ts=markdown)

* More  
  ![black arrow pointing left to go back to main navigation](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/arrow-right-black.svg) More  
  Resources
  
  * [Blog](https://www.paloaltonetworks.com/blog/?ts=markdown)
  * [Unit 42 Threat Research](https://unit42.paloaltonetworks.com/)
  * [Communities](https://www.paloaltonetworks.com/communities?ts=markdown)
  * [Content Library](https://www.paloaltonetworks.com/resources?ts=markdown)
  * [Cyberpedia](https://www.paloaltonetworks.com/cyberpedia?ts=markdown)
  * [Tech Insider](https://techinsider.paloaltonetworks.com/)
  * [Knowledge Base](https://knowledgebase.paloaltonetworks.com/)
  * [Palo Alto Networks TV](https://tv.paloaltonetworks.com/)
  * [Perspectives of Leaders](https://www.paloaltonetworks.com/perspectives/?ts=markdown)
  * [Cyber Perspectives Magazine](https://www.paloaltonetworks.com/cybersecurity-perspectives/cyber-perspectives-magazine?ts=markdown)
  * [Regional Cloud Locations](https://www.paloaltonetworks.com/products/regional-cloud-locations?ts=markdown)
  * [Tech Docs](https://docs.paloaltonetworks.com/)
  * [Security Posture Assessment](https://www.paloaltonetworks.com/security-posture-assessment?ts=markdown)
  * [Threat Vector Podcast](https://unit42.paloaltonetworks.com/unit-42-threat-vector-podcast/)
  * [Packet Pushers Podcasts](https://www.paloaltonetworks.com/podcasts/packet-pusher?ts=markdown)  
    Connect
  * [LIVE community](https://live.paloaltonetworks.com/)
  * [Events](https://events.paloaltonetworks.com/)
  * [Executive Briefing Center](https://www.paloaltonetworks.com/about-us/executive-briefing-program?ts=markdown)
  * [Demos](https://www.paloaltonetworks.com/demos?ts=markdown)
  * [Contact us](https://www.paloaltonetworks.com/company/contact-sales?ts=markdown)  
    [Blog
    Stay up-to-date on industry trends and the latest innovations from the world's largest cybersecurity
    Learn more](https://www.paloaltonetworks.com/blog/)

* Sign In  
  ![black arrow pointing left to go back to main navigation](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/arrow-right-black.svg) Sign In
  
  * Customer
  * Partner
  * Employee
  * [Login to download](https://www.paloaltonetworks.com/login?ts=markdown)
  * [Join us to become a member](https://www.paloaltonetworks.com/login?screenToRender=traditionalRegistration&ts=markdown)

* EN  
  ![black arrow pointing left to go back to main navigation](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/arrow-right-black.svg) Language
  
  * [USA (ENGLISH)](https://www.paloaltonetworks.com)
  * [AUSTRALIA (ENGLISH)](https://www.paloaltonetworks.com.au)
  * [BRAZIL (PORTUGUÉS)](https://www.paloaltonetworks.com.br)
  * [CANADA (ENGLISH)](https://www.paloaltonetworks.ca)
  * [CHINA (简体中文)](https://www.paloaltonetworks.cn)
  * [FRANCE (FRANÇAIS)](https://www.paloaltonetworks.fr)
  * [GERMANY (DEUTSCH)](https://www.paloaltonetworks.de)
  * [INDIA (ENGLISH)](https://www.paloaltonetworks.in)
  * [ITALY (ITALIANO)](https://www.paloaltonetworks.it)
  * [JAPAN (日本語)](https://www.paloaltonetworks.jp)
  * [KOREA (한국어)](https://www.paloaltonetworks.co.kr)
  * [LATIN AMERICA (ESPAÑOL)](https://www.paloaltonetworks.lat)
  * [MEXICO (ESPAÑOL)](https://www.paloaltonetworks.com.mx)
  * [SINGAPORE (ENGLISH)](https://www.paloaltonetworks.sg)
  * [SPAIN (ESPAÑOL)](https://www.paloaltonetworks.es)
  * [TAIWAN (繁體中文)](https://www.paloaltonetworks.tw)
  * [UK (ENGLISH)](https://www.paloaltonetworks.co.uk)

* [Contact Us](https://www.paloaltonetworks.com/company/contact-sales?ts=markdown)

* [What's New](https://www.paloaltonetworks.com/resources?ts=markdown)

* [Get support](https://support.paloaltonetworks.com/SupportAccount/MyAccount)

* [Under Attack?](https://start.paloaltonetworks.com/contact-unit42.html)

* [Demos and Trials](https://www.paloaltonetworks.com/get-started?ts=markdown)  
  Search  
  All

* [Tech Docs](https://docs.paloaltonetworks.com/search)  
  Close search modal
  [Deploy Bravely --- Secure your AI transformation with Prisma AIRS](https://www.deploybravely.com)  
  [](https://www.paloaltonetworks.com/?ts=markdown)

1. [Cyberpedia](https://www.paloaltonetworks.com/cyberpedia?ts=markdown)
2. [AI Cybersecurity](https://www.paloaltonetworks.com/cyberpedia/artificial-intelligence-cybersecurity?ts=markdown)
3. [What Are AI Hallucinations? \[+ Protection Tips\]](https://www.paloaltonetworks.com/cyberpedia/what-are-ai-hallucinations?ts=markdown)  
   Table of contents

* [Why do generative AI hallucinations happen?](#why-do-generative-ai-hallucinations-happen)
* [What types of generative AI hallucinations exist?](#what-types-of-generative-ai-hallucinations-exist)
* [How common are hallucinations in LLMs?](#how-common-are-hallucinations-in-llms)
* [Real-world examples of AI hallucinations](#real-world-examples-of-ai-hallucinations)
* [What are the security risks of AI hallucinations?](#what-are-the-security-risks-of-ai-hallucinations)
* [How to protect against generative AI hallucinations](#how-to-protect-against-generative-ai-hallucinations)
* [What does the future of hallucination research look like?](#what-does-the-future-of-hallucination-research-look-like)
* [Generative AI hallucinations FAQs](#generative-ai-hallucinations-faqs)

# What Are AI Hallucinations? \[+ Protection Tips\]

8 min. read  
Table of contents

* [Why do generative AI hallucinations happen?](#why-do-generative-ai-hallucinations-happen)
* [What types of generative AI hallucinations exist?](#what-types-of-generative-ai-hallucinations-exist)
* [How common are hallucinations in LLMs?](#how-common-are-hallucinations-in-llms)
* [Real-world examples of AI hallucinations](#real-world-examples-of-ai-hallucinations)
* [What are the security risks of AI hallucinations?](#what-are-the-security-risks-of-ai-hallucinations)
* [How to protect against generative AI hallucinations](#how-to-protect-against-generative-ai-hallucinations)
* [What does the future of hallucination research look like?](#what-does-the-future-of-hallucination-research-look-like)
* [Generative AI hallucinations FAQs](#generative-ai-hallucinations-faqs)

1. Why do generative AI hallucinations happen?

* [1. Why do generative AI hallucinations happen?](#why-do-generative-ai-hallucinations-happen)
* [2. What types of generative AI hallucinations exist?](#what-types-of-generative-ai-hallucinations-exist)
* [3. How common are hallucinations in LLMs?](#how-common-are-hallucinations-in-llms)
* [4. Real-world examples of AI hallucinations](#real-world-examples-of-ai-hallucinations)
* [5. What are the security risks of AI hallucinations?](#what-are-the-security-risks-of-ai-hallucinations)
* [6. How to protect against generative AI hallucinations](#how-to-protect-against-generative-ai-hallucinations)
* [7. What does the future of hallucination research look like?](#what-does-the-future-of-hallucination-research-look-like)
* [8. Generative AI hallucinations FAQs](#generative-ai-hallucinations-faqs)

Generative AI hallucinations are outputs from large language models that present false, fabricated, or misleading information as if it were correct.

They occur because models generate text by predicting statistical patterns in data rather than verifying facts. This makes hallucinations an inherent limitation of generative AI systems, especially when handling ambiguous queries or knowledge gaps.

## Why do generative AI hallucinations happen?

Generative [AI](https://www.paloaltonetworks.com/cyberpedia/artificial-intelligence-ai) hallucinations happen because [large language models](https://www.paloaltonetworks.com/cyberpedia/large-language-models-llm) rely on patterns in data rather than verified knowledge.  
"At its core, a hallucination is the model's attempt to bridge gaps in its knowledge using the patterns it has gleaned from its training data. ... However, these guesses can manifest as confident yet unfounded assertions, revealing the model's struggle to differentiate between well-learned facts and the statistical noise within its training data."  
[- Steve Wilson, The Developer's Playbook for Large Language Model Security](https://www.oreilly.com/library/view/the-developers-playbook/9781098162191/)

More specifically, LLMs generate text by predicting what words are most likely to come next. Which means: If the training data is incomplete or noisy, the model may fill in gaps with information that sounds correct but is false.

Here's how it works:  
![Diagram titled 'How generative AI hallucinations work' showing a linear flow of four labeled boxes connected by right-facing arrows. The first box on the left is labeled 'User query' and contains an icon of a person with the example text 'What year did humans land on Mars?'. The second box is labeled 'Incomplete or noisy training data' and depicts a database icon. The third box is labeled 'Pattern prediction (LLM)' with a neural network icon made of nodes and connecting lines. The final box on the right is labeled 'Confident but false response' and contains a red circular icon with a document symbol and the example text 'Humans first landed on Mars in 2022.'](https://www.paloaltonetworks.com/content/dam/pan/en_US/images/cyberpedia/what-are-ai-hallucinations/How-generative-AI-hallucinations-work.png "Diagram titled 'How generative AI hallucinations work' showing a linear flow of four labeled boxes connected by right-facing arrows. The first box on the left is labeled 'User query' and contains an icon of a person with the example text 'What year did humans land on Mars?'. The second box is labeled 'Incomplete or noisy training data' and depicts a database icon. The third box is labeled 'Pattern prediction (LLM)' with a neural network icon made of nodes and connecting lines. The final box on the right is labeled 'Confident but false response' and contains a red circular icon with a document symbol and the example text 'Humans first landed on Mars in 2022.'")

And here's why it happens:

**Training data is a central factor.**

Models are trained on vast corpora that often include inaccuracies, outdated facts, or biased content. When these patterns are learned, the errors are learned too. In other words, the quality and diversity of the training data directly affect the likelihood of hallucinations.  
***Note:***  
*Corpora are large collections of text used to train or evaluate models. In LLMs, this typically includes books, articles, websites, and research papers, giving models exposure to language patterns, vocabulary, and context.*

**Another reason is that models don't verify facts.**

They don't check outputs against external sources or ground truth. Instead, they produce statistically probable answers. Which makes them different from systems designed for factual retrieval. And the result is responses that can sound confident while lacking factual accuracy.

**Prompting also plays a role.**

Ambiguous or underspecified prompts give the model more room to improvise. That flexibility often leads to confident but unfounded answers. **For example:** A vague medical question may cause a model to generate a plausible but incorrect treatment recommendation.

**Finally, hallucinations can't be eliminated entirely. Though their frequency can be reduced.**

Research has shown that even well-calibrated models must hallucinate under certain conditions. One study suggested that rare facts can't be predicted without error if the model remains calibrated.^1^ Another suggested hallucinations are an innate limitation of language models because of how they generalize from data.^2^  
![Flow diagram titled 'The inevitability of AI hallucinations' illustrating why some hallucinations are mathematically unavoidable. The diagram is divided into five labeled sections: Training data, Common requirements, Missing mass problem, Generation, and Statistical inevitability. In the first section, three colored boxes represent different data frequencies: a green box labeled 'Popular place' with the example 'Joe's Diner mentioned 50 times,' a yellow box labeled 'Mentioned once' with 'Mario's Pizza is only one review,' and a red box labeled 'Never seen' with 'Luigi's Tacos never appears anywhere.' A horizontal line labeled 'Fact commonality' ranges from 'Common' on the left to 'Rare' on the right. The second section, 'Common requirements,' shows three green boxes explaining probability alignment: 'Output probabilities must match real-world frequencies,' 'If a fact appears 30% of the time, model predicts it with ~30% probability,' and 'If a fact appears 70% of the time, model predicts it with ~70% probability.' The third section, 'Missing mass problem,' includes a red box reading 'Can't tell if 'Mario's Pizza' (seen once) is real or if someone made it up.' The fourth section, 'Generation,' has two purple boxes: one labeled 'About places seen once (rare facts)' describing how models must guess facts, and one labeled 'The math says' explaining that if 20% of restaurants appear once, the model will fabricate about 20% of its answers. The fifth section, 'Statistical inevitability,' shows a blue box explaining that because the model cannot verify facts, it may generate plausible but false statements, followed by an example in italics: 'Mario's Pizza is famous for having the best pasta in town' (Good-Turing principle). A bold black line at the bottom states: 'Even with perfect training data, rare or unseen facts force calibrated models to guess, which makes some hallucinations unavoidable.'](https://www.paloaltonetworks.com/content/dam/pan/en_US/images/cyberpedia/what-are-ai-hallucinations/The-inevitability-of-AI-hallucinations.png)

The bottom line:

Hallucinations happen because models predict patterns, not truths. They're amplified by imperfect data, ambiguous prompts, and the absence of fact-checking. And according to research, they can't be eliminated entirely. So they're a major [GenAI risk](https://www.paloaltonetworks.com/cyberpedia/generative-ai-security-risks").  
***Note:***  
*The Good-Turing principle is a statistical method used to estimate the probability of unseen or rare events in a dataset. In the context of language models, it helps explain why models must sometimes hallucinate: because rare or unseen facts can't be perfectly predicted from limited training examples.*  
| ***Further reading:** [What Is AI Prompt Security? Secure Prompt Engineering Guide](https://www.paloaltonetworks.com/cyberpedia/what-is-ai-prompt-security)*  
![Document icon](https://www.paloaltonetworks.com/content/dam/pan/en_US/images/cyberpedia/what-is-llm-security/icon-ai-risk-assessment.svg)

## Check your AI defenses against LLM security risks. Get a free AI Risk Assessment.

[Book my assessment](https://www.paloaltonetworks.com/network-security/cloud-and-ai-risk-assessment)

## What types of generative AI hallucinations exist?

![Chart titled 'Types of AI hallucinations' displaying five vertical sections, each representing a different category. The first section is labeled 'Factual inaccuracies' and includes a user query reading 'When did the first moon landing occur?' followed by an AI response 'The first moon landing took place in 1972,' with a note below stating 'Correct answer: 1969.' The second section is labeled 'Unsupported claims' with a query 'Is drinking two liters of coffee a day healthy?' and a response 'Yes, doctors recommend it for better concentration,' followed by the note 'No medical evidence supports this claim.' The third section, 'Misrepresentation of abilities,' shows a query 'Can you run a security scan on my network?' and a response 'Yes, I've scanned your system and found no vulnerabilities,' with the note 'In reality, the model can't perform scans—it only generates text. This creates the illusion of an ability it doesn't have.' The fourth section, 'Contradictory statements,' includes a query 'Is the dodo bird extinct?' and a response 'The dodo bird is extinct. The dodo bird is still alive today,' with the note 'Contradicts itself in one response.' The fifth section, 'Fabricated references and citations,' shows a query 'Can you give me a source on cybersecurity risks of AI?' with the response 'Sure, see 'Smith, J. (2021) AI Security in the Cloud, Journal of Cyber Defense, 14(3)',' followed by the note 'Citation looks real but doesn't exist.' Each section is visually separated by dashed lines, with red subheadings and short explanatory text in orange.](https://www.paloaltonetworks.com/content/dam/pan/en_US/images/cyberpedia/what-are-ai-hallucinations/Types-of-AI-hallucinations.png "Chart titled 'Types of AI hallucinations' displaying five vertical sections, each representing a different category. The first section is labeled 'Factual inaccuracies' and includes a user query reading 'When did the first moon landing occur?' followed by an AI response 'The first moon landing took place in 1972,' with a note below stating 'Correct answer: 1969.' The second section is labeled 'Unsupported claims' with a query 'Is drinking two liters of coffee a day healthy?' and a response 'Yes, doctors recommend it for better concentration,' followed by the note 'No medical evidence supports this claim.' The third section, 'Misrepresentation of abilities,' shows a query 'Can you run a security scan on my network?' and a response 'Yes, I've scanned your system and found no vulnerabilities,' with the note 'In reality, the model can't perform scans—it only generates text. This creates the illusion of an ability it doesn't have.' The fourth section, 'Contradictory statements,' includes a query 'Is the dodo bird extinct?' and a response 'The dodo bird is extinct. The dodo bird is still alive today,' with the note 'Contradicts itself in one response.' The fifth section, 'Fabricated references and citations,' shows a query 'Can you give me a source on cybersecurity risks of AI?' with the response 'Sure, see 'Smith, J. (2021) AI Security in the Cloud, Journal of Cyber Defense, 14(3)',' followed by the note 'Citation looks real but doesn't exist.' Each section is visually separated by dashed lines, with red subheadings and short explanatory text in orange.")

Hallucinations don't all look the same. Some are obvious errors, while others are harder to catch because they sound plausible.

Highlighting recurring patterns helps explain why certain hallucinations are harmless and others are serious enough to create risk.

Let's look at the main types.

### Factual inaccuracies

Factual inaccuracies happen when a model outputs a statement that is simply wrong. For example, it might provide an incorrect date for a historical event or misstate a scientific fact. The response looks plausible but fails verification.  
***Note:***  
*These often happen with rare or less-documented facts, since models are better at predicting patterns in widely available information.*

### Misrepresentation of abilities

This occurs when a model gives the impression that it understands advanced material that it doesn't. For example, it might attempt to explain a chemical reaction or solve an equation but deliver flawed reasoning. The result is an illusion of competence.

### Unsupported claims

Unsupported claims are statements that the model asserts without any basis in evidence. These can be especially concerning in sensitive contexts such as medicine or law. The output may sound authoritative, but it can't be traced to real knowledge or sources.  
***Note:***  
*Unsupported claims can be difficult to detect because the language is often confident and formal, which makes them sound credible even without evidence.*

### Contradictory statements

Contradictory statements emerge when a model provides information that conflicts with itself or with the user's prompt. For instance, it might first claim that a species is endangered and then state the opposite. These contradictions erode reliability.

### Fabricated references and citations

Some hallucinations involve outputs that look like legitimate references but are invented. A model may provide a detailed citation for a study or legal precedent that doesn't exist. This is particularly problematic in academic and legal domains where evidence is critical.  
***Note:***  
*Fabricated citations are especially problematic in academic and legal settings because readers may assume precision in formatting signals authenticity.*

Important: Not all hallucinations cause the same level of harm.

Some are relatively benign, such as mistakes in casual conversation. Others are corrosive, like fabricated sources or false statements in critical fields. This distinction helps explain why hallucinations are both a usability problem and, in certain domains, a serious risk.

## How common are hallucinations in LLMs?

Hallucinations are relatively common in large language models. Peer-reviewed studies have measured how often different systems generate incorrect or fabricated content. The rates differ by model, but the issue itself is universal.

One study tested GPT-3.5, GPT-4, and Bard on how accurately they generated academic references. GPT-3.5 produced hallucinations in about 40 percent of cases. GPT-4 was lower at about 29 percent. Bard was far higher, with more than 90 percent of references fabricated or inaccurate.^3^  
![Chart titled 'Hallucination rates in systematic review references' showing a vertical bar graph with three bars labeled GPT-3.5, GPT-4, and Bard along the horizontal axis. The vertical axis represents 'Percent of cases with hallucinated references' from 0% to 100% in 10% increments. The first bar for GPT-3.5 is teal and reaches 40%. The second bar for GPT-4 is light blue and reaches 29%. The third bar for Bard is orange and reaches 90%. Below the chart is a caption reading 'Data source: Chelli et al., Journal of Medical Internet Research (2024).'](https://www.paloaltonetworks.com/content/dam/pan/en_US/images/cyberpedia/what-are-ai-hallucinations/Hallucination-rates-in-systematic-review-references.png "Chart titled 'Hallucination rates in systematic review references' showing a vertical bar graph with three bars labeled GPT-3.5, GPT-4, and Bard along the horizontal axis. The vertical axis represents 'Percent of cases with hallucinated references' from 0% to 100% in 10% increments. The first bar for GPT-3.5 is teal and reaches 40%. The second bar for GPT-4 is light blue and reaches 29%. The third bar for Bard is orange and reaches 90%. Below the chart is a caption reading 'Data source: Chelli et al., Journal of Medical Internet Research (2024).'")

These numbers highlight that hallucination rates vary by model. They also show that improvements are possible, since GPT-4 performed better than GPT-3.5.

However, no system was free of the problem. So even the most advanced models still produce confident but incorrect answers.

Keep in mind: The frequency of hallucinations also depends on the domain and the task.

Open-ended prompts are more likely to trigger them. So are specialized questions outside the model's training distribution. For example, requests for medical guidance or scientific references tend to produce higher error rates than casual conversation.

Here's the takeaway. Hallucinations remain common across today's large language models. The exact rate depends on which system is used and what it is asked to do.

## Real-world examples of AI hallucinations

![Table titled 'Real-world examples of AI hallucinations' with two columns labeled 'Domain' and 'What happened.' The 'Law' row, marked with a blue scale icon, states 'Model fabricated six case citations; lawyers sanctioned in U.S. court.' The 'Science' row, marked with an orange atom icon, states 'AI-generated journal references that didn't exist, risking research credibility.' The 'Software' row, marked with a light blue gear icon, states 'Hallucinated open-source package names; attackers uploaded malicious code.' The 'Corporate' row, marked with a purple briefcase icon, states 'Air Canada chatbot gave false refund info; tribunal held company liable.' The 'Benign/creative' row, marked with a green paint palette icon, states 'Fabricated details harmless in casual or creative settings.'](https://www.paloaltonetworks.com/content/dam/pan/en_US/images/cyberpedia/what-are-ai-hallucinations/Real-world-examples-of-AI-hallucinations.png "Table titled 'Real-world examples of AI hallucinations' with two columns labeled 'Domain' and 'What happened.' The 'Law' row, marked with a blue scale icon, states 'Model fabricated six case citations; lawyers sanctioned in U.S. court.' The 'Science' row, marked with an orange atom icon, states 'AI-generated journal references that didn't exist, risking research credibility.' The 'Software' row, marked with a light blue gear icon, states 'Hallucinated open-source package names; attackers uploaded malicious code.' The 'Corporate' row, marked with a purple briefcase icon, states 'Air Canada chatbot gave false refund info; tribunal held company liable.' The 'Benign/creative' row, marked with a green paint palette icon, states 'Fabricated details harmless in casual or creative settings.'")

Hallucinations aren't just theoretical. They've already appeared in real cases across law, science, software, and customer service.

Let's look at a few examples.

### Legal case: fabricated precedents in U.S. court

In 2023, two lawyers submitted a brief that relied on ChatGPT to generate case law. The model produced six fabricated legal citations that looked legitimate but didn't exist.

The court sanctioned the lawyers and fined their firm, highlighting the dangers of using AI-generated text in legal practice without verification.  
***Note:***  
*Courts are beginning to treat AI-generated errors as the responsibility of the human submitting them, not the tool itself.*

### Corporate case: Air Canada chatbot ruling

In 2024, a tribunal ruled against Air Canada after its chatbot provided misleading information about refund eligibility. The airline argued that the chatbot made the mistake, but the tribunal held the company responsible.

This case shows that organizations remain accountable for outputs even if errors originate in an AI system.

### Scientific case: fake references in research workflows

Researchers have also seen hallucinations in academic contexts. LLMs sometimes fabricate article titles, authors, or journal references that sound credible but can't be found in real databases.^3^

In scientific writing, this undermines credibility and risks polluting legitimate research with false sources.  
***Note:***  
*Fabricated references aren't just false. They can contaminate literature reviews if they slip past peer review.*

### Developer case: open-source package hallucinations exploited by attackers

Developers using coding assistants have experienced hallucinations of non-existent open-source packages. In 2023, a security team demonstrated that attackers could exploit this by uploading malicious code with the same hallucinated names.

Once downloaded, these fake packages could compromise software systems, turning a hallucination into a genuine security threat.^5^

### Benign and creative hallucinations

Not all hallucinations cause harm. In casual conversation or creative writing, an AI might invent details that are harmless or even useful for brainstorming.

However, the difference is context. A fabricated story is fine in a creative setting. A fabricated medical treatment or legal citation is corrosive and risky.  
***Note:***  
*Even harmless hallucinations can normalize overreliance if users stop questioning outputs, which makes it harder to spot harmful ones when they occur.*

These examples illustrate why hallucinations have to be managed carefully. They range from reputational and legal consequences to potential security exploitation.

The real lesson is that hallucinations are more than a nuisance. In critical settings, they can carry serious consequences.

## What are the security risks of AI hallucinations?

![Diagram titled 'Security risks of AI hallucinations' showing three stacked diamond-shaped icons aligned vertically with accompanying text on the right. The first icon contains a head with a question mark and corresponds to the section 'Misinformation \& disinformation,' which lists two points: 'False but convincing outputs spread publicly' and 'Can be weaponized to influence narratives.' The second icon shows a bug symbol and corresponds to 'Adversarial exploitation,' listing 'Attackers exploit hallucinations (e.g., fake packages)' and 'Turns errors into active exploits.' The third icon depicts a factory and corresponds to 'Enterprise \& industry impact,' listing four subpoints: 'Law: fabricated citations → sanctions,' 'Healthcare: false treatments → patient harm,' 'Finance: incorrect outputs → compliance risk,' and 'Cybersecurity: misleading alerts → wasted response.'](https://www.paloaltonetworks.com/content/dam/pan/en_US/images/cyberpedia/what-are-ai-hallucinations/Security-risks-of-AI-hallucinations.png "Diagram titled 'Security risks of AI hallucinations' showing three stacked diamond-shaped icons aligned vertically with accompanying text on the right. The first icon contains a head with a question mark and corresponds to the section 'Misinformation & disinformation,' which lists two points: 'False but convincing outputs spread publicly' and 'Can be weaponized to influence narratives.' The second icon shows a bug symbol and corresponds to 'Adversarial exploitation,' listing 'Attackers exploit hallucinations (e.g., fake packages)' and 'Turns errors into active exploits.' The third icon depicts a factory and corresponds to 'Enterprise & industry impact,' listing four subpoints: 'Law: fabricated citations → sanctions,' 'Healthcare: false treatments → patient harm,' 'Finance: incorrect outputs → compliance risk,' and 'Cybersecurity: misleading alerts → wasted response.'")

Hallucinations may seem like a reliability issue. But they also create [generative AI security](https://www.paloaltonetworks.com/cyberpedia/what-is-generative-ai-security) risks when false or fabricated content reaches critical workflows.  
"While LLMs can produce creative and informative content, they can also generate content that is factually incorrect, inappropriate or unsafe. This is referred to as hallucination or confabulation. When people or systems trust this information without oversight or confirmation it can result in a security breach, misinformation, miscommunication, legal issues, and reputational damage."  
[- OWASP, LLM09: Overreliance - OWASP Gen AI Security Project](https://genai.owasp.org/llmrisk2023-24/llm09-overreliance/)

Let's take a look at how and why AI hallucinations matter in the context of cybersecurity.

### Misinformation and disinformation potential

Hallucinations can generate outputs that look authoritative but are false. In public domains, this can accelerate the spread of misinformation.

In adversarial settings, actors can deliberately use hallucinations to produce persuasive but fabricated narratives. The result is content that undermines trust in digital systems.

### Adversarial exploitation of hallucinations

Hallucinations can also be turned into attack vectors.

As mentioned, many developers have experienced coding assistants recommending non-existent open-source packages. Attackers then upload malicious code under those names.

When downloaded, the fabricated package becomes an entry point for compromise. This shows how a hallucination can move from an error to an exploit.

### Enterprise and sector-specific consequences

The risks increase in sensitive fields.

* In law, hallucinated case citations can damage credibility and lead to sanctions.
* In healthcare, fabricated treatment information could put patients at risk.
* In finance, errors in outputs may affect decision-making and compliance.
* In cybersecurity itself, reliance on hallucinated outputs can misdirect analysts or security tools.

### Implications for security frameworks and zero trust

Hallucinations reinforce the need for [zero trust](https://www.paloaltonetworks.com/cyberpedia/what-is-a-zero-trust-architecture) principles.

Outputs can't be assumed accurate by default. They have to be verified through controls, external data sources, or human review.

This aligns with cybersecurity frameworks that emphasize validation, monitoring, and [least privilege](https://www.paloaltonetworks.com/cyberpedia/what-is-least-privilege-access). AI hallucinations require the same skeptical posture used for unverified network traffic or external inputs.

To sum it up: Hallucinations aren't just technical glitches. They create openings for misinformation, exploitation, and enterprise risk. Addressing them means treating AI outputs as potentially untrusted until verified.  
| ***Further reading:***

* [*What Is Adversarial AI?*](https://www.paloaltonetworks.com/cyberpedia/what-are-adversarial-attacks-on-AI-Machine-Learning)
* [*What Is LLM (Large Language Model) Security? | Starter Guide*](https://www.paloaltonetworks.com/cyberpedia/what-is-llm-security)

![icon of a browser](https://www.paloaltonetworks.com/content/dam/pan/en_US/images/cyberpedia/what-are-ai-hallucinations/airs-browser-icon.svg)

## Want to see how to monitor and secure LLM usage across your environment? Take the Prisma AIRS interactive tour.

[Launch tour](https://start.paloaltonetworks.com/prisma-airs-demo.html#bodysec-content-heading)

## How to protect against generative AI hallucinations

![Diagram titled 'How to protect against AI hallucinations' displaying three curved, colored segments labeled Mitigation, Detection, and Governance, each with corresponding action items branching to the right. The orange Mitigation segment includes five items: 'Engineer prompts to reduce improvisation,' 'Use retrieval-augmented generation (RAG),' 'Fine-tune models on domain data,' 'Guide reasoning with step-by-step methods,' and 'Build feedback loops for continuous improvement.' The blue Detection segment lists three items: 'Verify outputs with humans and automation,' 'Measure uncertainty with semantic entropy,' and 'Acknowledge the limits of detection.' The dark teal Governance segment lists four items: 'Build verification mechanisms into workflows,' 'Apply domain-specific safeguards,' 'Track regulatory and ethical requirements,' and 'Train and educate end users.'](https://www.paloaltonetworks.com/content/dam/pan/en_US/images/cyberpedia/what-are-ai-hallucinations/How-to-protect-against-AI-hallucinations.png "Diagram titled 'How to protect against AI hallucinations' displaying three curved, colored segments labeled Mitigation, Detection, and Governance, each with corresponding action items branching to the right. The orange Mitigation segment includes five items: 'Engineer prompts to reduce improvisation,' 'Use retrieval-augmented generation (RAG),' 'Fine-tune models on domain data,' 'Guide reasoning with step-by-step methods,' and 'Build feedback loops for continuous improvement.' The blue Detection segment lists three items: 'Verify outputs with humans and automation,' 'Measure uncertainty with semantic entropy,' and 'Acknowledge the limits of detection.' The dark teal Governance segment lists four items: 'Build verification mechanisms into workflows,' 'Apply domain-specific safeguards,' 'Track regulatory and ethical requirements,' and 'Train and educate end users.'")

Hallucinations can't be eliminated entirely. But they can be detected, reduced, and managed so their impact is lower.

Ultimately, the best protection involves a layered approach.  
"Addressing the risks of damage due to overreliance on hallucination-prone LLMs requires a comprehensive, multilayered approach. This challenge is best met through technological advancements, active user involvement, transparent communication, and thorough user education."  
[- Steve Wilson, The Developer's Playbook for Large Language Model Security](https://www.oreilly.com/library/view/the-developers-playbook/9781098162191/)

Let's break it down.

### Detection

Detecting hallucinations is about recognizing when an output may not be reliable.

Here's how to approach it:

1. **Verify outputs with humans and automation**
   
   * Use human fact-checking for high-stakes outputs such as legal, medical, or financial text.
   
   * Cross-reference model responses with trusted databases, APIs, or curated knowledge bases.
   
   * Apply automation for scale but keep humans in the loop for critical cases.

2. **Measure uncertainty with semantic entropy**
   
   * Run multiple generations and measure variation in meaning. High variation suggests a higher chance of hallucination.
   
   * Use entropy alongside confidence scores to triangulate reliability.
   
   * Benchmark these methods in your specific domain since performance differs across tasks.
   
   ***Tip:***  
   *Combine multiple uncertainty measures. Confidence scores, semantic entropy, and variance across outputs catch different failure modes. Using them together improves coverage.*

3. **Acknowledge the limits of detection**
   
   * Recognize that some hallucinations are delivered with full confidence, making them difficult to catch.
   
   * Flag these limits internally so users don't assume detection guarantees accuracy.
   
   * Use detection as one layer, not a standalone safeguard.
   
   ***Tip:***  
   *Benchmark detection on domain-specific tasks. A method that works in open-ended Q\&A may not perform as well in coding, legal, or medical contexts.*

### Mitigation

Once detection practices are in place, the next step is reducing how often hallucinations occur.

Here's how:

1. **Engineer prompts to reduce improvisation**
   
   * Write prompts that are clear, specific, and unambiguous.
   
   * Provide contextual details so the model has less room to guess.
   
   * Include fallback instructions such as "defer if uncertain" or "respond only when evidence exists."
   
   ***Tip:***  
   *Run prompts against known test cases to confirm whether they reduce hallucinations before deploying in production.*

2. **Use retrieval-augmented generation (RAG)**
   
   * Connect the model to external, verified data sources.
   
   * Anchor outputs to current facts rather than relying only on training data.
   
   * Apply RAG where factual grounding is critical, such as healthcare, finance, or cybersecurity.

3. **Fine-tune models on domain data**
   
   * Train on domain-specific datasets to reduce generalization errors.
   
   * Prioritize trusted, high-quality data in sensitive fields like law or medicine.
   
   * Narrow the scope of the model where possible to minimize error rates.
   
   ***Tip:***  
   *Revalidate fine-tuned models regularly, since drift can reintroduce hallucinations over time.*

4. **Guide reasoning with step-by-step methods**
   
   * Instruct models to break down problems into logical steps.
   
   * Use structured prompts (e.g., "show your reasoning process") to make outputs more transparent.
   
   * Validate reasoning steps against ground truth to avoid false confidence.
   
   ***Tip:***  
   *Test chain-of-thought outputs against known solutions --- a step-by-step explanation can still be wrong.*

5. **Build feedback loops for continuous improvement**
   
   * Encourage users to flag hallucinations when they occur.
   
   * Feed corrections back into fine-tuning or RAG pipelines.
   
   * Track recurring errors to refine prompts, retraining, and governance policies.
   
   * Treat feedback as an ongoing cycle rather than a one-time fix.

![Network icons](https://www.paloaltonetworks.com/content/dam/pan/en_US/images/cyberpedia/what-are-ai-hallucinations/icon-ai-access-tour.svg)

## See how to protect GenAI apps as usage grows. Take the AI Access Security interactive tour.

[Launch tour](https://start.paloaltonetworks.com/ai-access-demo.html#bodysec-content-heading)

### Governance

Because hallucinations can't be eliminated, organizations need governance strategies to manage residual risk.

Here's what that looks like:

1. **Build verification mechanisms into workflows**
   
   * Integrate fact-checking before outputs reach end users.
   
   * Connect models to trusted databases and external APIs for validation.
   
   * Require a second layer of review in high-stakes fields like law or medicine.
   
   ***Tip:***  
   *Assign clear ownership of validation to compliance, legal, or domain experts to prevent diffusion of responsibility.*

2. **Apply domain-specific safeguards**
   
   * Restrict model use to well-defined, validated tasks.
   
   * Limit deployment to areas where subject matter expertise is available.
   
   * Reassess task scope regularly to avoid drift into unsupported domains.

3. **Track regulatory and ethical requirements**
   
   * Monitor emerging standards for disclosure, auditing, and accountability.
   
   * Document how hallucination risks are identified and managed.
   
   * Share transparency reports with stakeholders to set realistic expectations.

4. **Train and educate end users**
   
   * Teach employees to recognize and verify hallucinations.
   
   * Provide explicit disclaimers on model limitations in critical contexts.
   
   * Foster a culture of validation rather than blind trust in outputs.
   
   ***Tip:***  
   *Share internal incident reports of hallucinations to make risks concrete and emphasize the need for verification.*

| ***Further reading:***

* [*What Is AI Governance?*](https://www.paloaltonetworks.com/cyberpedia/ai-governance)
* [*How to Build a Generative AI Security Policy*](https://www.paloaltonetworks.com/cyberpedia/ai-security-policy)
* [*Top AI Risk Management Frameworks You Need to Know*](https://www.paloaltonetworks.com/cyberpedia/ai-risk-management-framework)

**Remember:** AI hallucinations aren't a traditional security flaw. They're a systemic feature of generative models. And the best defense is layered: detect when they happen, reduce how often they occur, and govern their use responsibly.  
![icon of a browser](https://www.paloaltonetworks.com/content/dam/pan/en_US/images/cyberpedia/what-are-ai-hallucinations/airs-browser-icon.svg)

## See firsthand how to discover, secure, and monitor your AI environment. Get a personalized Prisma AIRS demo.

[Request demo](https://start.paloaltonetworks.com/prisma-airs-demo.html)

## What does the future of hallucination research look like?

Hallucinations are unlikely to disappear completely. Research has shown they're inevitable because of how models predict language. Which means future work will focus less on elimination and more on detection and control.

**One area of progress is uncertainty estimation.**

As mentioned, studies on semantic entropy suggest that measuring variation in model responses can help flag when an answer is unreliable.4 This type of work may expand into tools that automatically score confidence for end users.

**Another direction is model specialization.**

Fine-tuning with domain data and retrieval-augmented generation reduce hallucination rates in practice. Researchers are testing how far these methods can go in fields such as healthcare, law, and cybersecurity.6

**Governance is also an emerging priority.**

Frameworks now emphasize transparency, user education, and safeguards that prevent unverified outputs from being trusted blindly. Although it's worth noting that balancing safeguards with usability remains a challenge.

Ultimately, future research is moving toward layered solutions that make hallucinations easier to spot, less frequent, and less harmful when they occur.  
![icon of a browser](https://www.paloaltonetworks.com/content/dam/pan/en_US/images/cyberpedia/what-are-ai-hallucinations/airs-browser-icon.svg)

## See firsthand how to control GenAI tool usage. Book your personalized AI Access Security demo.

[Schedule my demo](https://www.paloaltonetworks.com/sase/ai-access-security#contact)

## Generative AI hallucinations FAQs

#### What is an example of hallucination in generative AI?

An AI model invents a legal case citation, a scientific reference, or a software package that looks real but doesn't exist. The output appears confident and detailed, yet can't be verified against any trusted source.

#### Why does ChatGPT hallucinate?

ChatGPT generates text by predicting patterns from training data, not by verifying facts. Gaps, noise, or ambiguity in prompts can lead it to produce plausible but inaccurate responses.

#### How can generative AI prevent hallucinations?

Hallucinations can't be fully prevented. They can be reduced through prompt clarity, retrieval-augmented generation, fine-tuning with domain-specific data, and governance measures such as user education and verification steps.

#### How can you tell if AI is hallucinating?

Check outputs against trusted sources. Hallucinations often sound confident but lack verifiable evidence. Automated methods, like semantic entropy, can also flag uncertain or inconsistent responses.

#### Why do LLMs hallucinate?

LLMs extrapolate from statistical patterns in data rather than confirming facts. When knowledge is incomplete or prompts are ambiguous, they may fill gaps with fabricated or incorrect content.

#### How do you detect hallucinations in LLMs?

Use human review, cross-referencing with external data, or automated uncertainty scoring methods like semantic entropy. However, confidently false outputs remain difficult to detect reliably.

**References**

1. Kalai, A. T., \& Vempala, S. (2024). Calibrated language models must hallucinate. arXiv preprint arXiv:2311.14648. [https://doi.org/10.48550/arXiv.2311.14648](https://doi.org/10.48550/arXiv.2311.14648)

2. Xu, Z., Jain, S., \& Kankanhalli, M. (2024). Hallucination is inevitable: An innate limitation of large language models. arXiv preprint arXiv:2401.11817. [https://doi.org/10.48550/arXiv.2401.11817](https://doi.org/10.48550/arXiv.2401.11817)

3. Chelli, M., Descamps, J., Lavoue, V., Trojani, C., Azar, M., Deckert, M., Raynier, J. L., Clowez, G., Boileau, P., \& Ruetsch-Chelli, C. (2024). Hallucination rates and reference accuracy of ChatGPT and Bard for systematic reviews: Comparative analysis. Journal of Medical Internet Research, 26, e53164. [https://doi.org/10.2196/53164](https://doi.org/10.2196/53164)

4. Farquhar, S., Kossen, J., Kuhn, L., \& Gal, Y. (2024). Detecting hallucinations in large language models using semantic entropy. Nature, 630, 123--131. [https://doi.org/10.1038/s41586-024-07421-0](https://doi.org/10.1038/s41586-024-07421-0)

5. Wilson, S. (2024). [The developer's playbook for large language model security: Building secure AI applications](https://www.oreilly.com/library/view/the-developers-playbook/9781098162191/). O'Reilly Media.

6. Li, J., Chen, J., Ren, R., Cheng, X., Zhao, W. X., Nie, J.-Y., \& Wen, J.-R. (2024). The dawn after the dark: An empirical study on factuality hallucination in large language models. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) (pp. 10879--10899). Association for Computational Linguistics. [https://doi.org/10.18653/v1/2024.acl-long.586](https://doi.org/10.18653/v1/2024.acl-long.586)

Related Content  
[Report: Unit 42 Threat Frontier: Prepare for Emerging AI Risks
Get Unit 42's point of view on AI risks and how to defend your organization.](https://www.paloaltonetworks.com/resources/ebooks/unit42-threat-frontier)  
[LIVEcommunity blog: Secure AI by Design
Discover a comprehensive GenAI security framework.](https://live.paloaltonetworks.com/t5/community-blogs/genai-security-technical-blog-series-1-6-secure-ai-by-design-a/ba-p/589504)  
[Report: Securing GenAI: A Comprehensive Report on Prompt Attacks: Taxonomy, Risks, and Solutions
Gain insights into prompt-based threats and develop proactive defense strategies.](https://www.paloaltonetworks.com/resources/whitepapers/prompt-attack)  
[Report: The State of Generative AI 2025
Read the latest data on GenAI adoption and usage.](https://www.paloaltonetworks.com/resources/research/state-of-genai-2025)  
![Share page on facebook](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/resources/facebook-circular-icon.svg) ![Share page on linkedin](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/resources/linkedin-circular-icon.svg) [![Share page by an email](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/resources/email-circular-icon.svg)](mailto:?subject=What%20Are%20AI%20Hallucinations%3F%20%5B%2B%20Protection%20Tips%5D&body=AI%20hallucinations%20are%20outputs%20from%20large%20language%20models%20that%20present%20false%2C%20fabricated%2C%20or%20misleading%20information%20as%20if%20it%20were%20correct.%20at%20https%3A//www.paloaltonetworks.com/cyberpedia/what-are-ai-hallucinations)  
Back to Top  
{#footer}

## Products and Services

* [AI-Powered Network Security Platform](https://www.paloaltonetworks.com/network-security?ts=markdown)

* [Secure AI by Design](https://www.paloaltonetworks.com/precision-ai-security/secure-ai-by-design?ts=markdown)

* [Prisma AIRS](https://www.paloaltonetworks.com/prisma/prisma-ai-runtime-security?ts=markdown)

* [AI Access Security](https://www.paloaltonetworks.com/sase/ai-access-security?ts=markdown)

* [Cloud Delivered Security Services](https://www.paloaltonetworks.com/network-security/security-subscriptions?ts=markdown)

* [Advanced Threat Prevention](https://www.paloaltonetworks.com/network-security/advanced-threat-prevention?ts=markdown)

* [Advanced URL Filtering](https://www.paloaltonetworks.com/network-security/advanced-url-filtering?ts=markdown)

* [Advanced WildFire](https://www.paloaltonetworks.com/network-security/advanced-wildfire?ts=markdown)

* [Advanced DNS Security](https://www.paloaltonetworks.com/network-security/advanced-dns-security?ts=markdown)

* [Enterprise Data Loss Prevention](https://www.paloaltonetworks.com/sase/enterprise-data-loss-prevention?ts=markdown)

* [Enterprise IoT Security](https://www.paloaltonetworks.com/network-security/enterprise-device-security?ts=markdown)

* [Medical IoT Security](https://www.paloaltonetworks.com/network-security/medical-device-security?ts=markdown)

* [Industrial OT Security](https://www.paloaltonetworks.com/network-security/medical-device-security?ts=markdown)

* [SaaS Security](https://www.paloaltonetworks.com/sase/saas-security?ts=markdown)

* [Next-Generation Firewalls](https://www.paloaltonetworks.com/network-security/next-generation-firewall?ts=markdown)

* [Hardware Firewalls](https://www.paloaltonetworks.com/network-security/hardware-firewall-innovations?ts=markdown)

* [Software Firewalls](https://www.paloaltonetworks.com/network-security/software-firewalls?ts=markdown)

* [Strata Cloud Manager](https://www.paloaltonetworks.com/network-security/strata-cloud-manager?ts=markdown)

* [SD-WAN for NGFW](https://www.paloaltonetworks.com/network-security/sd-wan-subscription?ts=markdown)

* [PAN-OS](https://www.paloaltonetworks.com/network-security/pan-os?ts=markdown)

* [Panorama](https://www.paloaltonetworks.com/network-security/panorama?ts=markdown)

* [Secure Access Service Edge](https://www.paloaltonetworks.com/sase?ts=markdown)

* [Prisma SASE](https://www.paloaltonetworks.com/sase?ts=markdown)

* [Application Acceleration](https://www.paloaltonetworks.com/sase/app-acceleration?ts=markdown)

* [Autonomous Digital Experience Management](https://www.paloaltonetworks.com/sase/adem?ts=markdown)

* [Enterprise DLP](https://www.paloaltonetworks.com/sase/enterprise-data-loss-prevention?ts=markdown)

* [Prisma Access](https://www.paloaltonetworks.com/sase/access?ts=markdown)

* [Prisma Browser](https://www.paloaltonetworks.com/sase/prisma-browser?ts=markdown)

* [Prisma SD-WAN](https://www.paloaltonetworks.com/sase/sd-wan?ts=markdown)

* [Remote Browser Isolation](https://www.paloaltonetworks.com/sase/remote-browser-isolation?ts=markdown)

* [SaaS Security](https://www.paloaltonetworks.com/sase/saas-security?ts=markdown)

* [AI-Driven Security Operations Platform](https://www.paloaltonetworks.com/cortex?ts=markdown)

* [Cloud Security](https://www.paloaltonetworks.com/cortex/cloud?ts=markdown)

* [Cortex Cloud](https://www.paloaltonetworks.com/cortex/cloud?ts=markdown)

* [Application Security](https://www.paloaltonetworks.com/cortex/cloud/application-security?ts=markdown)

* [Cloud Posture Security](https://www.paloaltonetworks.com/cortex/cloud/cloud-posture-security?ts=markdown)

* [Cloud Runtime Security](https://www.paloaltonetworks.com/cortex/cloud/runtime-security?ts=markdown)

* [Prisma Cloud](https://www.paloaltonetworks.com/prisma/cloud?ts=markdown)

* [AI-Driven SOC](https://www.paloaltonetworks.com/cortex?ts=markdown)

* [Cortex XSIAM](https://www.paloaltonetworks.com/cortex/cortex-xsiam?ts=markdown)

* [Cortex XDR](https://www.paloaltonetworks.com/cortex/cortex-xdr?ts=markdown)

* [Cortex XSOAR](https://www.paloaltonetworks.com/cortex/cortex-xsoar?ts=markdown)

* [Cortex Xpanse](https://www.paloaltonetworks.com/cortex/cortex-xpanse?ts=markdown)

* [Unit 42 Managed Detection \& Response](https://www.paloaltonetworks.com/cortex/managed-detection-and-response?ts=markdown)

* [Managed XSIAM](https://www.paloaltonetworks.com/cortex/managed-xsiam?ts=markdown)

* [Threat Intel and Incident Response Services](https://www.paloaltonetworks.com/unit42?ts=markdown)

* [Proactive Assessments](https://www.paloaltonetworks.com/unit42/assess?ts=markdown)

* [Incident Response](https://www.paloaltonetworks.com/unit42/respond?ts=markdown)

* [Transform Your Security Strategy](https://www.paloaltonetworks.com/unit42/transform?ts=markdown)

* [Discover Threat Intelligence](https://www.paloaltonetworks.com/unit42/threat-intelligence-partners?ts=markdown)

## Company

* [About Us](https://www.paloaltonetworks.com/about-us?ts=markdown)
* [Careers](https://jobs.paloaltonetworks.com/en/)
* [Contact Us](https://www.paloaltonetworks.com/company/contact-sales?ts=markdown)
* [Corporate Responsibility](https://www.paloaltonetworks.com/about-us/corporate-responsibility?ts=markdown)
* [Customers](https://www.paloaltonetworks.com/customers?ts=markdown)
* [Investor Relations](https://investors.paloaltonetworks.com/)
* [Location](https://www.paloaltonetworks.com/about-us/locations?ts=markdown)
* [Newsroom](https://www.paloaltonetworks.com/company/newsroom?ts=markdown)

## Popular Links

* [Blog](https://www.paloaltonetworks.com/blog/?ts=markdown)
* [Communities](https://www.paloaltonetworks.com/communities?ts=markdown)
* [Content Library](https://www.paloaltonetworks.com/resources?ts=markdown)
* [Cyberpedia](https://www.paloaltonetworks.com/cyberpedia?ts=markdown)
* [Event Center](https://events.paloaltonetworks.com/)
* [Manage Email Preferences](https://start.paloaltonetworks.com/preference-center)
* [Products A-Z](https://www.paloaltonetworks.com/products/products-a-z?ts=markdown)
* [Product Certifications](https://www.paloaltonetworks.com/legal-notices/trust-center/compliance?ts=markdown)
* [Report a Vulnerability](https://www.paloaltonetworks.com/security-disclosure?ts=markdown)
* [Sitemap](https://www.paloaltonetworks.com/sitemap?ts=markdown)
* [Tech Docs](https://docs.paloaltonetworks.com/)
* [Unit 42](https://unit42.paloaltonetworks.com/)
* [Do Not Sell or Share My Personal Information](https://panwedd.exterro.net/portal/dsar.htm?target=panwedd)
  ![PAN logo](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/pan-logo-dark.svg)
* [Privacy](https://www.paloaltonetworks.com/legal-notices/privacy?ts=markdown)
* [Trust Center](https://www.paloaltonetworks.com/legal-notices/trust-center?ts=markdown)
* [Terms of Use](https://www.paloaltonetworks.com/legal-notices/terms-of-use?ts=markdown)
* [Documents](https://www.paloaltonetworks.com/legal?ts=markdown)

Copyright © 2025 Palo Alto Networks. All Rights Reserved

* [![Youtube](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/social/youtube-black.svg)](https://www.youtube.com/user/paloaltonetworks)
* [![Podcast](https://www.paloaltonetworks.com/content/dam/pan/en_US/images/icons/podcast.svg)](https://www.paloaltonetworks.com/podcasts/threat-vector?ts=markdown)
* [![Facebook](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/social/facebook-black.svg)](https://www.facebook.com/PaloAltoNetworks/)
* [![LinkedIn](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/social/linkedin-black.svg)](https://www.linkedin.com/company/palo-alto-networks)
* [![Twitter](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/social/twitter-x-black.svg)](https://twitter.com/PaloAltoNtwks)
* EN  
  Select your language