Topic · Authentication
DKIM header oversigning
DKIM header oversigning signs additional occurrences of selected header names to constrain header insertion.
How it works
Repeated names in h= can make newly prepended fields detectable during verification.
What can go wrong
Without suitable header coverage, an intermediary may add confusing duplicate fields to a signed message.
What to check or do
Review oversigning support with the sending platform and test tampering with non-production messages.
Practical scenario
A legitimate message contains one From header, but a malicious or faulty intermediate service prepends a second field that may confuse recipients. Oversigning selected headers can make that insertion detectable during DKIM verification, though it does not eliminate every header-abuse scenario.
Verification checklist
- Record the exact signed header list in DKIM h=.
- Check whether the sending platform supports oversigning critical fields.
- Test altered-header samples in a nonproduction environment.
- Compare raw receiver-side fields with the trusted original message.
How to investigate DKIM header oversigning
- Signal to investigate
- An attacker inserts a second unsigned From header into a loosely protected message.
- Evidence to collect
- Inspect h= and whether repeated security-sensitive header field names were signed defensively.
- Common wrong conclusion
- Signing one occurrence of a header is not always equivalent to protecting against added copies.
Put this into practice
Run a focused check, then explore protection if your domain needs it.