Skip to main content Home About the Design SystemRoadmap OverviewDesignersDevelopers OverviewColorGridIconographyInteractionsSpacingTypography Overview Global colorBox shadowTypographyBorderOpacitySpaceLengthIconBreakpointsMedia queries All elements Accordion Alert Announcement Audio player Avatar Back to top Badge Blockquote Breadcrumb Button group Button Card Chip Code block Call to action Dialog Disclosure Footer Health index Icon Jump links Menu dropdown Navigation link Navigation (primary) Navigation (secondary) Navigation (vertical) Pagination PopoverPlanned Progress stepper Readtime Scheme toggle Select Site status Skeleton Skip link Spinner Statistic Subnavigation Surface Switch Table Tabs Tag Tile Timestamp Tooltip Video embed OverviewColor PalettesCustomizingDevelopers All PatternsAccordionAlertCall to ActionCardFilterFormLink with iconLogo wallSearch barSticky bannerSticky cardTabsTagTile AI design principlesLegal requirementsTransparency noticesIconographyColorChatbot avatarsAnimation All Personalization PatternsAnnouncement FundamentalsColorAccessibility toolsAssistive technologiesCI/CDContentContributorsDesignDevelopmentManual testingResourcesScreen readers Design/code status Release notes Get support

Transparency notices

It should always be clear when and how AI is being usedTransparency components for AI-related featuresTransparency notices for AI-assisted featuresTransparency notices for virtual assistantsIndicating AI-generated contentIndicating AI-generated imagesIt should always be clear when and how AI is being usedTransparency components for AI-related featuresTransparency notices for AI-assisted featuresTransparency notices for virtual assistantsIndicating AI-generated contentIndicating AI-generated images

It should always be clear when and how AI is being used

User research has shown that users want to clearly see when an action that they will take involves AI. Err on the side of over-communicating about AI features.

  • Don't rely on just one form of indicator for AI-enabled experiences.
  • At minimum use one visual and one verbal indicator, including options like icons, "with AI" button text, animations, and text disclaimers.
  • Consider additional indicators for high-risk interactions. Consult with the AIA Reviewers during your AIA review process as needed.

These components are used to notify users when they are interacting with AI-related features. They also inform users that generated responses may contain inaccuracies and advise them not to provide personal information, as AI systems may retain data for future improvements.

  • Common UX practices suggest placing the icon before the disclosure helps users more easily notice and recognize the message.
  • Tooltips are not included in these messages.
  • The label and notification style may vary depending on the context in which the information appears.
  • Please refer to the guidance for more details on notification messages.

Inline alerts

Use inline alerts to surface AI disclosures directly within the page context.

  • Inline alert warning component with "[Feature name] uses AI. Check for mistakes." guidance
    Warning variant: notify users that a feature uses AI and that they should review AI-generated content prior to use.
  • Inline alert info component with "Important. This feature uses AI technology. Do not include any personal information..." guidance
    Info variant: advise users not to include personal or sensitive information, as AI systems may retain data.

Toast alerts

Use toast alerts for more prominent AI disclosures, such as when a feature is first accessed.

  • Toast alert warning component with "[Feature name] uses AI. Check for mistakes." guidance
    Warning variant: inform users that their information will be used to create an AI-generated summary and that output may not be up-to-date or without error.
  • Toast alert info component with "Important. This feature uses AI technology. Do not include any personal information..." guidance
    Info variant: remind users not to include personal information, and link to the Red Hat Privacy Statement for more information.

Transparency notices for AI-assisted features

Some AI-assisted features may warrant more than an icon and text label. In these cases, a text notice can be placed at the beginning of an experience.

  • The text for this notice may vary or be tailored to the content.
  • For external-facing AI features, refer to this guidance and work with your AIA Reviewers during the AIA review process.
  • Inline alert warning component shown above an AI case summary within a security application, with "Compliance case uses AI. Check for mistakes..." guidance
    A warning notice placed above AI-generated content, alerting users to check for mistakes.
  • Inline alert info component shown above an AI case summary within a security application, with "Important. This feature uses AI technology..." guidance
    An info notice placed above AI-related tools, reminding users not to include sensitive information.

Transparency notices for virtual assistants

  • Include a notice at the beginning of an experience stating, at a minimum: "This feature uses AI technology. Do not include any personal information or other sensitive information in your input." (This notice is shown before a user interacts with the virtual assistant and before any content is generated.)
  • Include a persistent notice under the 'i' icon: "Always review AI-generated content prior to use." (This is a persistent notice associated with the input/output box.)
  • For external-facing AI features, refer to this guidance.
  • Ask Red Hat virtual assistant with the AI icon in the header bar as a visual indicator, and the AI tag/label as a verbal indicator
    The AI icon in the header bar serves as a visual indicator, and the "AI" tag/label is a verbal indicator.
  • Ask Red Hat virtual assistant with an alert info component giving notice before the user interacts, with a persistent notice associated with the input/output box
    A notice shown before the user interacts with the virtual assistant, with a persistent notice associated with the input/output box.

Indicating AI-generated content

AI-generated content, like a search results summary, must include a label and an icon indicating to the user that the content has been created using AI.

Example: AI-assisted results expandable section with the title indicating the use of AI both visually, with the sparkle icon, and verbally, by including "AI-assisted" in the results heading.
The title indicates the use of AI both visually, with the sparkle icon, and verbally, by including "AI-assisted" in the results heading.

Indicating AI-generated images

AI-generated images or ads should not be labeled individually.

  • Visually indicating each image would be distracting and could cause users to assume that entire images or pages are created without human influence, as seen in the responses to our AI sentiment survey.
  • Instead, Red Hat is working on a site-wide notice that will outline when users will (and won't) encounter AI-generated images on Red Hat web properties.
Example: Desktop view of redhat.com showing an AI-generated laptop image, and Red Hat Developer ad with an AI-generated image of the "Repo" character with the birthday cake.

  • The laptop in this collage was generated with an Adobe Firefly custom model, but the rest of the elements and the page content are not genAI.

  • The "Repo" character and the birthday cake in this ad were both generated using AI tools, but the logos, copy, and composition were created and composed by humans.

Edit this page on GitHub
© 2026 Red Hat Deploys by Netlify