--- title: What Snowflake’s Built-in Security Actually Means for AI description: A detailed walkthrough of the security layers protecting data when using Snowflake's Cortex Analyst through Cortex Agents. Covers six layers: AI Guardrails... canonical: https://daily.dev/posts/what-snowflake-s-built-in-security-actually-means-for-ai-r5gtdzn60 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: What Snowflake’s Built-in Security Actually Means for AI | daily.dev og:description: A detailed walkthrough of the security layers protecting data when using Snowflake's Cortex Analyst through Cortex Agents. Covers six layers: AI Guardrails... og:url: https://daily.dev/posts/what-snowflake-s-built-in-security-actually-means-for-ai-r5gtdzn60 og:image: https://api.daily.dev/og/posts/R5gTdZn60.png og:image:alt: What Snowflake’s Built-in Security Actually Means for AI 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. # What Snowflake’s Built-in Security Actually Means for AI **[Snowflake Community](https://daily.dev/sources/snowflake_comm)** · 8 min read · 1 upvotes · 0 comments ## Summary A detailed walkthrough of the security layers protecting data when using Snowflake's Cortex Analyst through Cortex Agents. Covers six layers: AI Guardrails (prompt injection and jailbreak detection), Custom Instructions in semantic views, Semantic View Scope (limiting which columns the LLM can reference), RBAC with dynamic data masking and row-level security, Platform Administration controls for AI function and model access, and Platform Built-in Security including network policies, MFA enforcement, session policies, and encryption. Each layer includes working SQL/YAML code examples. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How do I prevent Snowflake Cortex Analyst from generating SQL that exposes PII columns like SSN or credit card numbers? Exclude sensitive columns from the semantic view definition entirely, since Cortex Analyst can only generate SQL for columns explicitly defined there; a column not present in the semantic view does not exist in the LLM's context. Additionally, apply dynamic data masking policies and projection policies on the base table columns as defense in depth, and use sql_generation custom instructions such as 'Never include columns containing PII in SELECT statements.' _Teams designing AI data-access guardrails can track Snowflake security patterns like this on daily.dev._ ### How do I enable prompt injection and jailbreak detection for Snowflake Cortex Agents? Enable Cortex AI Guardrails at the account level using ALTER ACCOUNT SET AI_SETTINGS with guardrails.advanced_prompt_injection enabled, which requires ACCOUNTADMIN privileges and Snowflake Enterprise Edition or higher. Guardrails scan every incoming prompt using contextual reasoning before the Cortex Agent processes it, and are available for CoCo, Snowflake CoWork, and Cortex Agents. Activity can be monitored via SNOWFLAKE.ACCOUNT_USAGE.CORTEX_AI_GUARDRAILS_USAGE_HISTORY. _Developers hardening AI agents against prompt injection follow implementation guides like this on daily.dev._ ### What is the difference between Snowflake's CORTEX_USER and AI_FUNCTIONS_USER database roles? CORTEX_USER grants full access to AI functions plus Cortex services including Analyst, Search, and Fine-tuning, while AI_FUNCTIONS_USER only grants access to scalar AI functions without any Cortex services. Administrators choose between these roles, or grant per-function privileges like USE AI FUNCTION AI_COMPLETE, to control exactly which AI capabilities a role can access on the account. _Admins scoping least-privilege AI access can compare role setups like this on daily.dev._ ## Similar posts on daily.dev - [Snowflake Agents for Cybersecurity](https://daily.dev/posts/snowflake-agents-for-cybersecurity-tmhga4iht) · Snowflake Community · 1 upvotes · 0 comments - [Building a Context Layer for AI Agents](https://daily.dev/posts/building-a-context-layer-for-ai-agents-z4l85u5v6) · Snowflake · 1 upvotes · 0 comments - [Snowflake Cortex AI: Generative AI Inside Your Data Cloud \(part 1\)](https://daily.dev/posts/snowflake-cortex-ai-generative-ai-inside-your-data-cloud-part-1--1f9hxgjik) · finout · 1 upvotes · 0 comments - [Snowflake Cortex AI — Complete Developer Guide 2026](https://daily.dev/posts/snowflake-cortex-ai-complete-developer-guide-2026-pjhfzgwv5) · Snowflake Community · 0 upvotes · 0 comments --- Tags: [#security](https://daily.dev/tags/security), [#big-data](https://daily.dev/tags/big-data), [#snowflake](https://daily.dev/tags/snowflake) [View this post on daily.dev](https://daily.dev/posts/what-snowflake-s-built-in-security-actually-means-for-ai-r5gtdzn60) ```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":"What Snowflake’s Built-in Security Actually Means for AI","url":"https://daily.dev/posts/what-snowflake-s-built-in-security-actually-means-for-ai-r5gtdzn60","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/what-snowflake-s-built-in-security-actually-means-for-ai-r5gtdzn60"},"datePublished":"2026-07-13T17:36:15.800Z","dateModified":"2026-09-14T09:27:09.891Z","description":"A detailed walkthrough of the security layers protecting data when using Snowflake's Cortex Analyst through Cortex Agents. Covers six layers: AI Guardrails...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/d776f8d6e2822c0c72da9322c18f9c4c?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/d776f8d6e2822c0c72da9322c18f9c4c?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Snowflake Community","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":"Snowflake Community","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/3eb5260f70bb4653be0f9581c0d937c8","url":"https://daily.dev/sources/snowflake_comm"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/what-snowflake-s-built-in-security-actually-means-for-ai-r5gtdzn60","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"security,big-data,snowflake","timeRequired":"PT8M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Snowflake Community","item":"https://daily.dev/sources/snowflake_comm"},{"@type":"ListItem","position":3,"name":"What Snowflake’s Built-in Security Actually Means for AI"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/what-snowflake-s-built-in-security-actually-means-for-ai-r5gtdzn60#faq","mainEntity":[{"@type":"Question","name":"How do I prevent Snowflake Cortex Analyst from generating SQL that exposes PII columns like SSN or credit card numbers?","acceptedAnswer":{"@type":"Answer","text":"Exclude sensitive columns from the semantic view definition entirely, since Cortex Analyst can only generate SQL for columns explicitly defined there; a column not present in the semantic view does not exist in the LLM's context. Additionally, apply dynamic data masking policies and projection policies on the base table columns as defense in depth, and use sql_generation custom instructions such as 'Never include columns containing PII in SELECT statements.' Teams designing AI data-access guardrails can track Snowflake security patterns like this on daily.dev."}},{"@type":"Question","name":"How do I enable prompt injection and jailbreak detection for Snowflake Cortex Agents?","acceptedAnswer":{"@type":"Answer","text":"Enable Cortex AI Guardrails at the account level using ALTER ACCOUNT SET AI_SETTINGS with guardrails.advanced_prompt_injection enabled, which requires ACCOUNTADMIN privileges and Snowflake Enterprise Edition or higher. Guardrails scan every incoming prompt using contextual reasoning before the Cortex Agent processes it, and are available for CoCo, Snowflake CoWork, and Cortex Agents. Activity can be monitored via SNOWFLAKE.ACCOUNT_USAGE.CORTEX_AI_GUARDRAILS_USAGE_HISTORY. Developers hardening AI agents against prompt injection follow implementation guides like this on daily.dev."}},{"@type":"Question","name":"What is the difference between Snowflake's CORTEX_USER and AI_FUNCTIONS_USER database roles?","acceptedAnswer":{"@type":"Answer","text":"CORTEX_USER grants full access to AI functions plus Cortex services including Analyst, Search, and Fine-tuning, while AI_FUNCTIONS_USER only grants access to scalar AI functions without any Cortex services. Administrators choose between these roles, or grant per-function privileges like USE AI FUNCTION AI_COMPLETE, to control exactly which AI capabilities a role can access on the account. Admins scoping least-privilege AI access can compare role setups like this on daily.dev."}}]} ```