Skip to main content
kcronin
Gainsight Employee ⭐️
Gainsight Employee ⭐️
May 13, 2026

Gainsight CC: Stop Spam Before It Starts with AI-Powered User Profile Moderation

Related products:CC Analytics & ReportingCC ModerationCC AI
  • May 13, 2026
  • 16 replies
  • 346 views

The Moderation AI Agent now screens new user registrations automatically, stopping spam bots and NSFW accounts before they ever set foot in your community.

 


Screen New User Profiles with AI-Powered Moderation

AI screening at the point of registration When a new account is created, the Moderation AI Agent evaluates it using the same intelligence applied to community posts. Accounts flagged for spam or NSFW content are automatically set to Banned and moved to the Banned Users group, with no manual review required.

Five profile fields analyzed The AI evaluates username, email address, signature, custom profile fields, and avatar image to make its determination.

Role-based exclusions built in Admins, Community Managers, Moderators, and Superusers are never screened, regardless of your settings.

Get Started Navigate to AI > Moderation Agent in Customer Communities Control and turn on the User profile moderation toggle.

 


AI Answers Analytics in the Search Analytics Dashboard (Beta)

If you're running AI Answers, you can now measure how members are actually engaging with it. The Search Analytics Dashboard has been updated with a dedicated AI Answers Analytics section, giving you clear visibility into answer adoption and quality.

New metrics now tracked:

  • Total Summaries Generated: How many AI Answers have been created

  • Impressions: Unique views or interactions with an AI Answer lasting 10+ seconds

  • Sources Clicked: Clicks on inline links and cited sources within AI Answers

  • Continue Reading Clicked: How often users expanded a full AI Answer

  • Show All Sources Clicked: How often users expanded the sources list

  • Helpful / Unhelpful Feedback: Direct thumbs-up and thumbs-down signals from members

Note: AI Answers Analytics metrics are measured starting from April 28, 2026.

To access the dashboard, navigate to Analytics > New Dashboards (BETA) in Customer Communities Control.

 


Make Images Clickable in Posts and Topics

You can now hyperlink images directly in the rich text editor. After uploading an image to a post or topic, click on it and select the Hyperlink option to attach a URL. Once published, clicking the image opens the linked URL in a new tab, turning any image into a clickable, navigable element. Images without a hyperlink will continue to open in a lightbox. Images with a hyperlink skip the lightbox and take the user directly to the linked URL instead.


👉 Check out the Release Notes for more details.
💬 Have questions or feedback? Drop a comment below — or better yet, ask our new AI!

 

    16 replies

    Daniele Cmty
    Helper ⭐️⭐️
    May 13, 2026

    These are amazing improvements, especially the Screen New User Profiles with AI-Powered Moderation is gonna be a game-changer for spam-prevention and wil address the issue at the core. thanks so much for that!

    Erik_
    Contributor ⭐️⭐️⭐️⭐️⭐️
    May 19, 2026

    Screen New User Profiles with AI-Powered Moderation

    AI screening at the point of registration When a new account is created, the Moderation AI Agent evaluates it using the same intelligence applied to community posts. Accounts flagged for spam or NSFW content are automatically set to Banned and moved to the Banned Users group, with no manual review required.

     

    We have an SSO sign-in and registration. Does this work then?

    The shortest answer is doing -George Herbert
    Daniele Cmty
    Helper ⭐️⭐️
    May 19, 2026


    Quick question:

    • where can we see the criteria that AI uses for Profile Moderation?
    • is it possible to personalize/alter these criteria or can we rely on the default system?



      Thanks in advance
    laura.porcincula
    Contributor ⭐️⭐️⭐️⭐️
    May 28, 2026

    So excited about image hyperlinking! How do I get it to open in a new tab? 

    Contributor ⭐️⭐️
    May 28, 2026

    +1 ^ 

    Contributor ⭐️⭐️
    June 2, 2026

    I really like the questions asked by ​@Daniele Cmty here!

    I was also wondering, some users that post spammy content in the community can have “regular” looking email addresses (with gmail etc) as well as “regular” usernames. On the contrary, users who genuinely want to engage can also have VERY strange usernames or email addresses. We wouldn’t want those to be mistakenly flagged. How would these cases be assessed by the AI user profile moderation tool?

    Thank you!

    Daniele Cmty
    Helper ⭐️⭐️
    June 2, 2026

    that should be a question for ​@kcronin  :)

    Larry
    Gainsight Employee ⭐️
    Gainsight Employee ⭐️
    June 2, 2026

    So excited about image hyperlinking! How do I get it to open in a new tab? 

    @laura.porcincula ​@maxpetesprout It should be opening in a new tab by default! Let me know if that’s not what you are experiencing? I’ll take a look, if so.

    Larry Imgrund || Senior Product Manager & Community Nerd
    Contributor ⭐️⭐️
    June 2, 2026

    For me, it’s not!

    Larry
    Gainsight Employee ⭐️
    Gainsight Employee ⭐️
    June 4, 2026

    I really like the questions asked by ​@Daniele Cmty here!

    I was also wondering, some users that post spammy content in the community can have “regular” looking email addresses (with gmail etc) as well as “regular” usernames. On the contrary, users who genuinely want to engage can also have VERY strange usernames or email addresses. We wouldn’t want those to be mistakenly flagged. How would these cases be assessed by the AI user profile moderation tool?

    Thank you!

    Great question, ​@zszwedo, that is the precise nuance we have to contend with.

    The agent isn't pattern-matching on whether an email or username *looks* unusual on its own. It evaluates the profile fields together (username, email, signature, custom fields, and avatar) rather than judging any single one in isolation. So a perfectly normal-looking Gmail address paired with a spammy signature can still get flagged, while a quirky username on an otherwise legitimate profile generally won't.

    No AI moderation system is perfect, and we're being deliberately careful about precision rather than over-promising a zero-false-positive rate. The important safety net: anything the agent bans lands in the Banned Users group rather than disappearing, so it's fully reviewable and reversible if something gets caught incorrectly.

    Larry Imgrund || Senior Product Manager & Community Nerd