Why does my email show a spoofing warning to people?
An email spoofing warning appears when the recipient's mail system cannot confirm that your domain really sent the message. In Outlook and Gmail that usually means SPF (Sender Policy Framework) or DKIM (DomainKeys Identified Mail) did not pass for a domain that aligns with your From address, so DMARC (Domain-based Message Authentication, Reporting and Conformance) failed. Fix the sending service, not the recipient.
The visible symptom differs per client: a question mark on the sender photo, a "via" tag next to the name, a banner, or delivery to Junk. Underneath, the cause is almost always one of three things: a sending service that is not authenticated for your domain, a service that authenticates with its own domain instead of yours, or a forwarding path that breaks authentication. This guide shows how to tell which, how to read the header evidence, and what an MSP (managed service provider) should tell the client. For the wider picture, start at the DMARC pillar guide.
Key takeaways
- Two different problems produce spoofing warnings: your own outbound mail failing authentication at the recipient, and inbound mail your users receive from senders that really are spoofed or simply unauthenticated. Diagnose which one you have before changing anything.
- Outlook's question mark appears when a message fails SPF and DKIM and also fails DMARC or Microsoft's composite authentication. The "via" tag appears when the DKIM domain or MAIL FROM domain differs from the From domain.
- Gmail shows a question mark next to the sender's name for unauthenticated mail and a "via" label when the sending domain does not match the From domain. Google recommends DKIM as the preferred fix.
- The fix is alignment: make each sending service sign with DKIM using your domain (
d=) or use a Return-Path on your domain, then confirmdmarc=passin real headers. - Do not tell clients to whitelist their way out. Allow entries hide the warning in one tenant and leave every other recipient seeing it.
Which warnings do Outlook and Gmail actually show?
Outlook shows a "?" in the sender photo, an underlined "via" tag, a first contact safety tip, an External tag and, in some cases, a spoof warning banner or Junk delivery. Gmail shows a question mark next to the sender name, a "via" label, and warning banners. The exact banner wording varies by client and version, so read the header rather than the banner.
What do the Microsoft 365 and Outlook indicators mean?
Microsoft documents these indicators in its anti-phishing policies reference. They are part of the spoof settings in anti-phishing policies, available for all cloud mailboxes (Exchange Online Protection) as well as Defender for Office 365.
- Question mark on the sender photo. The setting "Show (?) for unauthenticated senders for spoof" adds it when the message does not pass SPF or DKIM checks and does not pass DMARC or composite authentication. It needs spoof intelligence turned on. Microsoft notes that "for the question mark in the sender's photo, SPF or DKIM are the most important."
- "via" tag. Shown when the From domain differs from the domain in the DKIM signature or the MAIL FROM address. The fix is to make the DKIM domain or MAIL FROM domain match the From domain, or be a subdomain of it.
- First contact safety tip. Shows "You don't often get email from" followed by the address, the first time a recipient gets mail from a sender or when they rarely do. It is not an authentication failure. Microsoft's recommended settings list it as off by default and on in the Standard and Strict presets.
- External tag. Configured with
Set-ExternalInOutlook, which adds an External icon near the subject line of messages from outside the organisation. It describes where mail came from, not whether it authenticated. - Spoof warning banner or Junk delivery. Microsoft's anti-spoofing protection page says users "might see a spoof warning banner in Outlook, the '?' symbol next to the sender name, or the message delivered to the Junk Email folder" when partner mail fails authentication.
The question mark and the via tag are on by default. Microsoft's consumer article on phishing and suspicious behaviour in Outlook also notes that spoof intelligence can mark a sender as "unverified", and lists Outlook.com, new Outlook for Windows and Outlook 2021/2024 as showing these cues.
What do the Gmail warnings mean?
Gmail's help page on checking whether a message is authenticated says: "If you see a question mark next to the sender's name, the message isn't authenticated." Google adds that unauthenticated messages "aren't necessarily spam", and asks senders to authenticate with DKIM (preferred) or SPF.
The "via" label appears when "the domain it was sent from doesn't match the domain in the 'From:' address", or when mail reaches a Google Group from a domain with p=reject or p=quarantine. Google's fix list for senders is to cover vendors in SPF, sign with DKIM "associated with your domain", and make the From domain match the authenticating domain.
Gmail also shows a "This message may not have been sent by…" warning when a message claims to come from a Gmail address but lacks authentication data showing it was sent through Gmail. Common causes are a website form sending "as" a Gmail user, or automatic forwarding.
Google Workspace admins can add stricter banners. The advanced phishing and malware protection settings cover mail pretending to be from your domain and unauthenticated mail from any domain, with a choice of warning banner, spam folder or quarantine. Banners appear in Gmail on the web, not in third-party apps.
Apple Mail does not have published, verifiable documentation of equivalent authentication indicators, so this guide does not cover it.
Is the warning about mail you send or mail you receive?
Ask who sees the warning. If your client's customers or suppliers report it on mail from your client's domain, the problem is outbound authentication and you fix it in DNS and sending services. If your client's staff see it on mail they receive, it is an inbound filtering verdict about someone else's domain.
Outbound (your domain triggers warnings elsewhere). Typical causes are a newsletter tool, CRM, invoicing system, ticketing desk or website form that sends as the client's domain without DKIM on that domain. Each recipient system judges independently, so the warning shows at Microsoft and Google recipients alike until the sender is fixed.
Inbound (your users see warnings). There are three cases:
- Real spoofing or a lookalike domain. The warning is doing its job. Do not allow-list it. Our guide to the types of email spoofing shows how to tell exact-domain spoofing, lookalike domains and display-name tricks apart. If users receive lures that show their own address as the sender, see why you get emails from your own address.
- A legitimate sender with weak authentication. A partner with no DMARC record or a broken SPF record fails Microsoft's implicit checks. Tell the partner; allow only as a narrow, documented exception.
- Your client's own devices or apps. A scanner or SaaS tool sending as the client's domain into its own tenant triggers intra-org spoofing (
compauth=fail reason=6xx). Fix it with a proper connector or DKIM, not an allow entry.
How do you triage each warning?
Match the warning text and client to the likely cause, then confirm in the header before changing DNS. The table covers the patterns MSP service desks see most often. Header evidence always wins over the banner, because the same banner can come from several underlying results.
| Warning or indicator | Client | Likely cause | Fix |
|---|---|---|---|
| "?" on sender photo | Outlook, Outlook.com | SPF and DKIM failed or unaligned, DMARC or compauth failed | Enable DKIM with your domain on the sending service; correct SPF |
| "via" tag next to sender | Outlook | DKIM d= and MAIL FROM both use the vendor's domain |
Configure custom DKIM or custom Return-Path on your domain |
| "You don't often get email from…" | Outlook | First contact safety tip, sender is new to recipient | Nothing to fix; explain it to users |
| External icon near subject | Outlook | Set-ExternalInOutlook is enabled |
Nothing to fix; it marks origin, not authentication |
Spoof banner or Junk, compauth=fail reason=000 |
Microsoft 365 | DMARC failed and sender publishes quarantine or reject | Align the failing sending service |
Junk, compauth=fail reason=001 |
Microsoft 365 | No DMARC, weak SPF or p=none and implicit checks failed |
Publish SPF, DKIM and DMARC properly |
compauth=fail reason=601 |
Microsoft 365 | Own domain sent into own tenant without authentication | Authenticate the device or app; use a connector |
| Question mark next to sender name | Gmail | Message not authenticated | Sign with DKIM on your domain; fix SPF |
| "via" next to sender | Gmail | Sending domain differs from From domain, or Google Group with enforcing DMARC | Align DKIM or Return-Path with the From domain |
| "This message may not have been sent by…" | Gmail | Gmail address used as From without Gmail authentication | Send from your own domain, not a Gmail address |
How do you read the Authentication-Results header?
Open the header (Gmail: More > Show original; Outlook: message properties or the Message Header Analyzer) and find Authentication-Results. Compare smtp.mailfrom and header.d with header.from. If neither matches the From domain, DMARC fails even when SPF and DKIM both show pass.
Here is a typical outbound failure as seen by a Microsoft 365 recipient. A marketing platform sent as example.com but authenticated as itself:
Authentication-Results: spf=pass (sender IP is 198.51.100.20)
smtp.mailfrom=bounces.mailer.example.net; dkim=pass (signature was verified)
header.d=mailer.example.net; dmarc=fail action=none header.from=example.com;
compauth=fail reason=001
X-Forefront-Antispam-Report: ...;CAT:SPOOF;...
Read it field by field, using Microsoft's anti-spam message header reference:
spf=pass ... smtp.mailfrom=bounces.mailer.example.net: SPF passed, but for the vendor's envelope domain, not yours.dkim=pass ... header.d=mailer.example.net: DKIM passed, again for the vendor's domain.dmarc=fail action=none header.from=example.com: neither pass aligns with the From domain, so DMARC fails.action=nonebecause the domain is atp=none.compauth=fail reason=001: Microsoft's composite verdict failed implicitly, because there was no enforcing policy to apply.
The same message after the fix looks like this:
Authentication-Results: spf=pass (sender IP is 198.51.100.20)
smtp.mailfrom=bounce.example.com; dkim=pass (signature was verified)
header.d=example.com; dmarc=pass action=none header.from=example.com;
compauth=pass reason=100
The key compauth reason codes are 000 (DMARC failed with quarantine or reject), 001 (implicit failure, including p=none), 100 (aligned pass) and 6xx (intra-org spoofing). Microsoft stresses that a compauth failure "doesn't directly result in blocking a message"; spoof intelligence and policy decide the action. For the full code table and policy behaviour, see DMARC in Microsoft Defender for Office 365. For Received chains and ARC, see our MSP guide to reading email headers.
How does an MSP fix spoofing warnings on a client's outbound mail?
Find the exact service that sent the flagged message, make that service authenticate with the client's domain, then prove the fix in a fresh header. The warning disappears for new mail once SPF or DKIM passes in alignment with the From domain. Work service by service; a domain-wide record change rarely fixes it alone.
- Collect one flagged message. Ask the recipient to forward it as an attachment, or ask for a screenshot plus the full header. Paste the header into the email header analyzer.
- Identify the sending service. The
smtp.mailfromdomain, theheader.ddomain, the sending IP and the Received chain normally name the platform. Your DMARC aggregate reports show the same source at volume. Our guide to finding every sending source covers the discovery method. - Align DKIM first. In the platform, enable custom domain authentication so it signs with
d=set to the client's domain or a subdomain of it, and publish the CNAME or TXT records it gives you. DKIM survives most forwarding, which is why Google calls it preferred. Platform steps are in how to set up DKIM; check the result with the DKIM validator. - Align the Return-Path where the platform allows it. A custom bounce domain such as
bounce.example.commakessmtp.mailfromalign, which gives SPF a chance to count for DMARC. Add the platform's include to SPF only if it sends with your domain in MAIL FROM, and stay within the ten-lookup limit (check with the SPF analyzer). - Publish or repair DMARC. Use one valid record at
_dmarc, starting atp=nonewith an aggregate reporting address so you can see every source:
_dmarc.example.com. IN TXT "v=DMARC1; p=none; rua=mailto:dmarc-reports(at)example.com"
Replace (at) with the at sign when you publish. Confirm the record with the DMARC checker. Under RFC 9989, which obsoletes RFC 7489, the old percentage tag is gone. Move from p=none to p=quarantine to p=reject once reports show all legitimate sources aligned, as described in moving from monitoring to enforcement safely. You can use t=y to signal a testing phase, and np= to set policy for non-existent subdomains.
- Retest. Send a new message from the fixed service to a Microsoft 365 and a Gmail mailbox you control. Check for
dkim=pass header.d=example.com,dmarc=passand, at Microsoft,compauth=pass. Old messages keep their original verdict, so always test with new mail.
If DMARC still fails with SPF passing, the envelope domain is not aligned; why DMARC fails when SPF passes explains the mechanics, and DMARC alignment explained covers relaxed versus strict modes.
What if the message was forwarded?
Forwarding changes the sending IP, so SPF breaks, and mailing lists that edit content can break DKIM. If the header shows a forwarder or list in the Received chain, the original sender may be fine. DKIM alignment gives the best chance of surviving; see why forwarded mail fails DMARC.
Should you add an allow entry instead?
Only as a last, documented resort for inbound mail. Microsoft's spoof intelligence and Tenant Allow/Block List can suppress the question mark and via tag for a sender pair, but only inside that one tenant. It does nothing for the client's other recipients, and it weakens protection against real spoofing of that domain.
How should an MSP handle this across a client portfolio?
Treat spoofing warnings as an onboarding and change-control signal. A new SaaS tool sending as the client's domain is the usual trigger, so add "how does it authenticate?" to every new-tool request. Track each client's sending services, DKIM selectors and DMARC policy centrally, and review aggregate reports for new unaligned sources.
For UK clients, the NCSC's email security and anti-spoofing guidance sets out the same order: deploy SPF and DKIM, publish DMARC, then enforce. Third-party platforms deserve a named owner on the client side; our guide to handling third-party senders covers the conversation with vendors. The DMARC for MSPs page describes how DMARC AI supports portfolio monitoring.
What should you tell the client?
Keep it short and factual: what recipients see, why, what you are changing and when to expect a difference. Avoid blaming the recipient's filter. Clients mainly want to know whether mail is being lost and whether they need to do anything.
A template you can adapt:
Subject: Why some recipients see a warning on your emails
Some recipients see a "?" or "via" label, or a security warning, on emails sent from [service name] on your domain. This happens because that service currently sends on your behalf without proving it is allowed to use your domain. It is not a sign that your mailbox has been hacked.
We are adding the authentication records [service name] provides to your DNS and switching on signing with your domain. This needs no action from your staff. New emails should stop showing the warning once the change is live; emails already sent will keep the old label.
If anyone in your team signs up for a new tool that sends email as your company, please let us know first so we can set it up correctly.
For inbound cases, the message is different: tell users the warning is about the sender's domain, ask them to verify unexpected payment or login requests by phone, and log the sender so you can contact the partner's IT team.
FAQ
Why does Outlook show a question mark on emails from my domain?
Outlook adds the question mark when the message fails SPF and DKIM and also fails DMARC or Microsoft's composite authentication. It usually means a service sending as your domain is not authenticated with it. Enable DKIM signing with your domain on that service, correct SPF, and check new headers for dmarc=pass.
How do I remove "via" from my emails in Gmail and Outlook?
Make the domain that authenticates the message match your From domain. In practice, enable custom DKIM on the sending platform so d= is your domain, or configure a custom Return-Path on your domain. Gmail users cannot remove the label, and an Outlook admin can only hide it inside their own tenant.
Does DMARC p=reject stop spoofing warnings?
No. A strict policy tells receivers what to do with mail that fails, but your own legitimate mail only stops triggering warnings once it passes aligned SPF or DKIM. Fix alignment for every sending service first, then move through p=quarantine to p=reject so attackers' mail is rejected instead of warned about.
Why do my users see warnings on mail from a trusted partner?
The partner's domain probably lacks a DMARC record, has a broken SPF record, or uses a service that is not aligned. Microsoft then fails the message implicitly, shown as compauth=fail reason=001. Ask the partner's IT team to fix authentication; use a narrow, documented allow entry only while they do.
Is the first contact safety tip a spoofing warning?
No. "You don't often get email from" only means the recipient rarely or never receives mail from that address. It appears regardless of authentication results and is designed to make users pause on unfamiliar senders. Explain it to users rather than trying to suppress it, because it helps against impersonation.