--- title: Shadow AI: Risks, Detection, and Security Strategies description: Shadow AI now extends beyond browser-tab chat tools to standalone agents on user devices and unregistered model endpoints inside cloud accounts. Unlike shadow... canonical: https://daily.dev/posts/shadow-ai-risks-detection-and-security-strategies-nge3aomif twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Shadow AI: Risks, Detection, and Security Strategies | daily.dev og:description: Shadow AI now extends beyond browser-tab chat tools to standalone agents on user devices and unregistered model endpoints inside cloud accounts. Unlike shadow... og:url: https://daily.dev/posts/shadow-ai-risks-detection-and-security-strategies-nge3aomif og:image: https://api.daily.dev/og/posts/NgE3aOMiF.png og:image:alt: Shadow AI: Risks, Detection, and Security Strategies og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Shadow AI: Risks, Detection, and Security Strategies **[Orca Security Blog](https://daily.dev/sources/orca-security-blog)** · 16 min read · 0 upvotes · 0 comments ## Summary Shadow AI now extends beyond browser-tab chat tools to standalone agents on user devices and unregistered model endpoints inside cloud accounts. Unlike shadow IT, revoking access to shadow AI cannot retract data already sent to a provider, so disposition must track data classes, retention, and training terms. A defensible inventory merges signals from identity providers, SaaS admin consoles, and network egress logs, since no single source covers the whole estate. Findings need four dispositions (sanction, sanction with conditions, restrict, block) with a named owner and service-level target to avoid piling up undecided. Frameworks like ISO/IEC 42001:2023 and the CSA AI Controls Matrix v1.1 give the resulting policy an auditable structure. Orca's cloud-side scanning covers AI inventory in cloud accounts (Azure OpenAI, Bedrock, SageMaker, Vertex AI) but not desktop agents or personal-account browser sessions. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How is shadow AI different from shadow IT in terms of remediation? Shadow IT is an access problem you can close by revoking access and reclaiming the stored copy; shadow AI is a disclosure problem you can only bound. Once a prompt reaches an AI provider, its retention and training terms decide what happens to that data next, and revoking access afterward only stops future access, not the data already sent. _Anyone building an AI governance policy can track these distinctions on daily.dev alongside other security posture discussions._ ### Can Microsoft Entra admins revoke OAuth consent grants for an AI app from the admin portal? No, grants made through user consent cannot be revoked in the Entra admin portal; removing a grant there does not stop users from re-consenting to the same permissions. Revoking those consent grants requires using the Microsoft Graph API or PowerShell instead. _Developers managing OAuth grants and identity risk can follow tooling changes like this on daily.dev._ ### What does Microsoft's shadow AI detection in Microsoft 365 actually block versus detect? Microsoft's shadow AI preview feature detects seven named agents including ChatGPT Desktop, Ollama Desktop, and Claude Desktop, but blocking is available for only one of them. Blocking applies only to managed Windows devices enrolled with Microsoft Intune, and enforcement can take anywhere from 15 minutes up to 8 hours to take effect. _Security teams evaluating endpoint AI controls can track feature rollouts like this on daily.dev._ ## Similar posts on daily.dev - [Why Shadow AI Is The New Shadow API](https://daily.dev/posts/why-shadow-ai-is-the-new-shadow-api-yf1d00gr0) · Nordic APIs · 0 upvotes · 0 comments - [Shadow AI : How to deal with unauthorized models and uncontrolled agents](https://daily.dev/posts/shadow-ai-how-to-deal-with-unauthorized-models-and-uncontrolled-agents-5q0xjcnbp) · SD Times · 1 upvotes · 0 comments - [How to Detect and Eliminate Shadow AI in 5 Steps](https://daily.dev/posts/how-to-detect-and-eliminate-shadow-ai-in-5-steps-amd6akroj) · JFrog · 0 upvotes · 0 comments - [Shadow AI is already inside your company. Here’s how to get control of it](https://daily.dev/posts/shadow-ai-is-already-inside-your-company-here-s-how-to-get-control-of-it-44m1i7iyo) · The Next Web · 1 upvotes · 0 comments - [Shadow AI: Agentic Access and the New Frontier of Data Risk](https://daily.dev/posts/shadow-ai-agentic-access-and-the-new-frontier-of-data-risk-4j85krplu) · Security Boulevard · 1 upvotes · 0 comments --- Tags: [#security](https://daily.dev/tags/security), [#oauth](https://daily.dev/tags/oauth) [View this post on daily.dev](https://daily.dev/posts/shadow-ai-risks-detection-and-security-strategies-nge3aomif) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Shadow AI: Risks, Detection, and Security Strategies","url":"https://daily.dev/posts/shadow-ai-risks-detection-and-security-strategies-nge3aomif","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/shadow-ai-risks-detection-and-security-strategies-nge3aomif"},"datePublished":"2026-08-13T14:24:32.812Z","dateModified":"2026-09-14T08:56:51.735Z","description":"Shadow AI now extends beyond browser-tab chat tools to standalone agents on user devices and unregistered model endpoints inside cloud accounts. Unlike shadow...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/2ae8e63c4b9e59310e5d424274a16deb?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/2ae8e63c4b9e59310e5d424274a16deb?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Orca Security Blog","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Orca Security Blog","logo":"https://media.daily.dev/image/upload/s--kkQFNboJ--/f_auto,q_auto/v1780213281/logos/orca-security-blog?_a=BAMAMiWQ0","url":"https://daily.dev/sources/orca-security-blog"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/shadow-ai-risks-detection-and-security-strategies-nge3aomif","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"security,oauth","timeRequired":"PT16M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Orca Security Blog","item":"https://daily.dev/sources/orca-security-blog"},{"@type":"ListItem","position":3,"name":"Shadow AI: Risks, Detection, and Security Strategies"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/shadow-ai-risks-detection-and-security-strategies-nge3aomif#faq","mainEntity":[{"@type":"Question","name":"How is shadow AI different from shadow IT in terms of remediation?","acceptedAnswer":{"@type":"Answer","text":"Shadow IT is an access problem you can close by revoking access and reclaiming the stored copy; shadow AI is a disclosure problem you can only bound. Once a prompt reaches an AI provider, its retention and training terms decide what happens to that data next, and revoking access afterward only stops future access, not the data already sent. Anyone building an AI governance policy can track these distinctions on daily.dev alongside other security posture discussions."}},{"@type":"Question","name":"Can Microsoft Entra admins revoke OAuth consent grants for an AI app from the admin portal?","acceptedAnswer":{"@type":"Answer","text":"No, grants made through user consent cannot be revoked in the Entra admin portal; removing a grant there does not stop users from re-consenting to the same permissions. Revoking those consent grants requires using the Microsoft Graph API or PowerShell instead. Developers managing OAuth grants and identity risk can follow tooling changes like this on daily.dev."}},{"@type":"Question","name":"What does Microsoft's shadow AI detection in Microsoft 365 actually block versus detect?","acceptedAnswer":{"@type":"Answer","text":"Microsoft's shadow AI preview feature detects seven named agents including ChatGPT Desktop, Ollama Desktop, and Claude Desktop, but blocking is available for only one of them. Blocking applies only to managed Windows devices enrolled with Microsoft Intune, and enforcement can take anywhere from 15 minutes up to 8 hours to take effect. Security teams evaluating endpoint AI controls can track feature rollouts like this on daily.dev."}}]} ```