Looking for a quick answer? Quick answers lists the questions customers ask most, each with a one-line answer.
What it is
Smart Banners adds a short warning banner at the top of emails that show signs of phishing or impersonation, inside each user's own Microsoft 365 or Google Workspace mailbox. The banner says in plain words what looks wrong, so the user can stop before clicking, replying or paying, and lets them Report email or Mark as safe in one click.
Smart Banners never blocks, quarantines or deletes an email: it only adds a warning, and the user decides what to do. Nothing is installed on users' devices: the banner is written into the email itself. Administrators activate users, manage safe lists and follow activity from Smart Banners in the LibraCyber console.
A Very suspicious banner: someone writes as a colleague from an address outside your company
How it works
Smart Banners reads each new email through your provider's API, looks for warning signs and, when it finds any, writes a banner at the top of the email.
Smart Banners works inside the mailbox, after your email filtering. An email that your email gateway, Microsoft 365 or Google Workspace stops or quarantines never reaches the mailbox, so Smart Banners never sees it.
- A new email arrives in the mailbox of an activated user. Microsoft 365 or Google Workspace tells Smart Banners straight away.
- Smart Banners reads the email through the provider's API, with the access an administrator granted at setup, and looks for warning signs in who sent it, how it was sent, its links and its attachments.
- It compares the sender with what it knows: your colleagues, your company's domains, the people this mailbox already exchanges email with, well-known brands, and your safe lists. A sender who is trusted gets no banner at all.
- If it finds warning signs, it writes a banner at the top of the email. The colour follows the most serious sign, and the banner lists every sign found. Weak signs alone, such as a different reply address, never produce a banner.
- The user decides: Report email sends the email to your security team in Threats; Mark as safe stops banners on future emails from that sender address, for that user only.
When a mailbox is activated for the first time, Smart Banners reads the past 52 weeks of sent and received email, once, so that people the user already writes to are not flagged as new. Until this reading is finished, emails in that mailbox do not get the first-time-sender sign. Emails already in the mailbox do not get a banner.
Microsoft 365 and Google Workspace
Microsoft 365 updates each email in place; Google Workspace holds new email until it is analysed, then replaces it with a copy.
| Microsoft 365 | Google Workspace | |
|---|---|---|
| Access | Microsoft Graph API, with consent given by a Microsoft 365 administrator | Gmail API, through a service account with domain-wide delegation set up in the Google Admin console |
| What happens to the email | The banner is written into the email itself. Every analysed email is updated, even without a banner, and plain-text emails are shown as formatted text from then on | New email waits under the label Mantra/Processing until it is analysed. Every analysed email is then replaced by a copy, carrying the banner when there is one, and the original goes to Trash. Emails with attachments that need no banner are released unchanged |
| Junk or Spam | Analysed: emails in Junk Email can get a banner | Not analysed |
| Emails larger than 10 MB | Analysed | Not analysed, no banner |
What users see in Outlook and Gmail, step by step: Banners and user actions › What happens to the email in the mailbox? Setup for each provider: Setup › Step 2 — API Access Setup.
Mail apps and languages
Because the banner is part of the email, users see it in the mail apps they already use. These include Outlook (desktop, web and mobile apps), Gmail (web and mobile app) and other mail apps connected to the mailbox. The one exception is Apple Mail on macOS with Google Workspace (see What is not covered). The banner is written in the user's language; 14 languages are supported, listed in Banners and user actions › Which language is the banner in?
How long it takes
Banners are added email by email as messages arrive, usually within seconds of delivery; nothing is recalculated afterwards. On Google Workspace, where new emails wait for the analysis before reaching the inbox, most appear within about 10 seconds and 95% within about a minute. If the analysis fails, the email is released without a banner after a few minutes.
Banner types
- Informative (grey, sometimes called "informational"): only mild signs, such as a sender writing to the user for the first time.
- Suspicious (orange): the sender resembles a colleague, your company, a known brand or a business partner, or an attachment is a type often used in attacks.
- Very suspicious (red, sometimes called "highly suspicious"): the sender's name exactly matches a colleague's, or the domain is your company's with a different ending, while the email comes from elsewhere.
- Report button only (optional): a single Report email button on emails that get no other banner.
- Test banner (green): appears on the setup test email to confirm that banners work.
Full description of each banner, its texts, its buttons and its languages: Banners and user actions › When does each banner appear?
What is not covered
- Users who are not active: only users with the status Active in Users get banners; Inactive users, users in Access error and users who Can't be activated get none. What each status means: Users › What do the user statuses mean, and how do I fix them?
- Mailboxes on Exchange Server or any other on-premises mail server, or on a provider other than Microsoft 365 and Google Workspace: not supported. In a hybrid setup, only the mailboxes hosted in Microsoft 365 can get banners.
- Users removed or deactivated in your directory: not listed, no banners.
- Microsoft 365 users without a licence: not listed and not activated, unless LibraCyber support enables them for your company on request.
- Microsoft 365 shared mailboxes: listed and analysed by default. They get banners without the Report email and Mark as safe buttons.
- Apple Mail on macOS with Google Workspace: not supported. The mailbox shows Can't be activated in Users. No other mail app is blocked.
- Emails Smart Banners does not analyse: calendar invitations, LibraCyber phishing simulation emails, encrypted emails, emails with no readable text and Microsoft 365 delivery failure reports. On Google Workspace, also emails larger than 10 MB, emails in Spam and emails with a file type Gmail blocks.
- Emails that arrived before activation, and emails moved into a Microsoft 365 mailbox more than 7 days after they were received (for example during a mailbox migration).
- Emails the user sends: never analysed.
- Trusted senders, which get no banner: senders on your company safe list → Safe list. Also senders the user marked as safe, people the user regularly exchanges email with, the official addresses of well-known brands and LibraCyber's own emails. Your colleagues too, if Trust internal colleagues is on → Preferences › Settings and controls.
- Widely used legitimate domains: LibraCyber keeps a list of them, and their emails skip the checks for imitation of your company and of well-known brands.
Data and AI
Details on data handling and AI are in the Service Description & Security, available on request.