How to Protect Against Email Phishing: SPF, DKIM and DMARC
The controls that actually reduce email phishing are SPF, DKIM and DMARC on your own domain, phishing-resistant multi-factor authentication, and a payment process that does not depend on an email instruction. The detail that matters: SPF checks the envelope sender, not the From line the recipient reads, so SPF and DKIM alone do not stop somebody spoofing your domain convincingly.
Topic: Cyber Threats, Scams & Online Safety in Cybersecurity & Ethical Hacking · Tutorial
Quick answer: The controls that actually reduce email phishing are SPF, DKIM and DMARC on your own domain, phishing-resistant multi-factor authentication, and a payment process that does not depend on an email instruction. The detail that matters: SPF checks the envelope sender, not the From line the recipient reads, so SPF and DKIM alone do not stop somebody spoofing your domain convincingly. DMARC alignment is what closes that gap. In India, an organisation has 6 hours to report a listed incident to CERT-In, and an individual who has lost money should call 1930 and file on the portal within 24 hours.
Key takeaways
- SPF authorises a sending host for the
HELOandMAIL FROMidentities. The recipient never sees either of those. - DKIM proves a domain took responsibility for a message. RFC 6376 says its signing identity need not match the From header at all.
- DMARC is the control that ties authentication to the visible From address, and it is the only one of the three that does.
- A DMARC record at
p=noneblocks nothing. It is a reporting mode, and many domains never leave it. - CERT-In requires listed incidents reported within 6 hours. Calling 1930 is not a complaint: the portal filing must follow within 24 hours.
Topics covered in this guide: why phishing still works, SPF, DKIM and DMARC and what each does and does not prove, how to check your own domain from the command line, multi-factor authentication, payment fraud procedure, recognising modern phishing, awareness training, reporting deadlines in India, and the order to do all of it in.
Table of Contents
- Why Does Phishing Still Work?
- What Do SPF, DKIM and DMARC Each Do?
- Why Are SPF and DKIM Not Enough on Their Own?
- How Do You Check Your Own Domain?
- Why Is Multi-Factor Authentication So Important?
- What Procedures Stop Payment Fraud?
- How Do You Recognise Modern Phishing?
- What Makes Awareness Training Effective?
- What Should Someone Do If They Click?
- What Must You Report, and How Quickly?
- What Is the Priority Order?
- References
Why Does Phishing Still Work?
Because it targets the decision rather than the system. A message that looks routine, arrives from a plausible sender and asks for something ordinary does not trigger suspicion, and the person acting on it is doing their job rather than making a mistake.
The useful consequence is that controls which depend on people noticing something will always have a failure rate, so the controls that matter are the ones that work when somebody does not notice. Those are the three protocols below, phishing-resistant authentication, and a payment process that an email cannot override.
What Do SPF, DKIM and DMARC Each Do?
These three are usually described as a set, which hides the fact that they answer three different questions.
| Question it answers | DNS record | Specification | |
|---|---|---|---|
| SPF | Is this host allowed to send for this envelope domain? | TXT | RFC 7208, April 2014 |
| DKIM | Has a domain cryptographically taken responsibility for this message? | TXT, at a selector | RFC 6376, September 2011 |
| DMARC | Does either of the above line up with the From address the reader sees, and what should happen if not? | TXT, at _dmarc | RFC 7489, March 2015 |
SPF publishes which hosts may send mail for a domain. RFC 7208 defines the check precisely: "An SPF check tests the authorization of a client host to emit mail with a given identity", and the identities are HELO and MAIL FROM. The RFC also settles a long-running confusion: "SPF records MUST be published as a DNS TXT (type 16) Resource Record (RR) [RFC1035] only." There is no SPF record type to use.
The ending of an SPF record carries most of its meaning. -all is a fail, "an explicit statement that the client is not authorized to use the domain in the given identity". ~all is a softfail, "a weak statement by the publishing ADMD that the host is probably not authorized", which receivers are not expected to reject on.
DKIM signs the message so a recipient can verify it came from a domain holding the matching key. RFC 6376 describes this as the signing domain claiming "some responsibility for a message by associating the domain with the message". Three limits in the same RFC are worth knowing: verification "asserts nothing else about protecting the end-to-end integrity of the message" beyond the hash, "signature verification failure does not force rejection of the message", and "no attempt is made to include encryption as part of the mechanism". DKIM is an accountability mechanism, not a confidentiality one.
DMARC is the policy layer. RFC 7489 calls it "a scalable mechanism by which a mail-originating organization can express domain-level policies and preferences".
Why Are SPF and DKIM Not Enough on Their Own?
This is the part most articles skip, and it is the whole reason DMARC exists.
A recipient reads the From: header. SPF never looks at it: RFC 7208 says the identities checked are HELO and MAIL FROM, both of which live in the SMTP envelope and are invisible in a normal mail client. DKIM does not necessarily look at it either: RFC 6376 states that the signing identity is not required to match an address in any particular header field.
So an attacker can send a message that passes SPF for a domain they control, carries a valid DKIM signature for that same domain, and still displays [email protected] to the person reading it. Both checks pass. The mail is still a forgery of your brand.
DMARC closes this by requiring alignment. A message passes DMARC when at least one of SPF or DKIM passes and the domain it passed for lines up with the domain in the From header. Alignment can be strict, requiring an exact match, or relaxed, requiring the same organisational domain.
The policy then says what a receiver should do with failures, and RFC 7489 defines three values:
p=none– "The Domain Owner requests no specific action be taken regarding delivery of messages." Nothing is blocked. This is a monitoring mode.p=quarantine– the owner "wishes to have email that fails the DMARC mechanism check be treated by Mail Receivers as suspicious", typically the spam folder.p=reject– the owner "wishes for Mail Receivers to reject email that fails the DMARC mechanism check".
A domain sitting at p=none has DMARC in the same sense that a domain with a disconnected alarm has an alarm. Publishing p=none first is correct, because the reports tell you which of your own legitimate senders would break. Staying there is the mistake.
How Do You Check Your Own Domain?
Three commands tell you where you stand. Run them against your own domain before reading any vendor report about it.
# SPF: a TXT record at the domain itself, starting v=spf1
dig TXT example.com +short | grep spf1
# DMARC: a TXT record at the _dmarc subdomain
dig TXT _dmarc.example.com +short
# DKIM: a TXT record at selector._domainkey, and you need the selector name
dig TXT selector1._domainkey.example.com +short
On Windows without dig:
nslookup -type=TXT example.com
nslookup -type=TXT _dmarc.example.com
How to read what comes back:
- No SPF record. Anyone can send as your envelope domain without contradiction.
- SPF ending
~all. A softfail, which RFC 7208 says receivers should not reject on. Move to-allonce you are confident the record is complete. - No
_dmarcrecord. Nothing ties authentication to the visible From address, so your domain can be spoofed convincingly. p=nonewith norua=. The worst of both: no enforcement and nobody receiving the reports that would justify enforcement.- DKIM selector unknown. Look at the
DKIM-Signatureheader of a message you sent; thes=tag is the selector.
Finding the DKIM selector from a received message:
# show the signing domain and selector from a saved message
grep -i '^DKIM-Signature' message.eml | tr ';' '\n' | grep -E '^\s*[ds]='
Why Is Multi-Factor Authentication So Important?
Because it changes what a stolen password is worth. Credential phishing is the most common objective, and a password on its own stops being enough.
The qualification that matters in 2026: not all second factors resist phishing. A one-time code typed into a fake page is captured and replayed by the attacker in real time, and that is now routine rather than advanced. Factors bound to the site's origin, such as hardware security keys and passkeys, cannot be replayed onto the real site because the browser will not release them to the wrong domain.
Our guides to multi-factor authentication protocols and why MFA is not always enough cover where each type fails.
What Procedures Stop Payment Fraud?
A rule that no payment instruction or bank-detail change is actioned on the strength of an email, however convincing, and however senior the apparent sender.
- Verify out of band, on a number already held, never one supplied in the message.
- Treat a change of bank details as a new vendor, with the same checks.
- Require a second approver above a threshold, and make it genuinely independent.
- Remove urgency as an override. Invented urgency is the mechanism of the attack, so a process that bends under pressure has no control at all.
This is the control that works when every technical measure has failed, because it does not depend on detecting the message.
How Do You Recognise Modern Phishing?
The old advice about spelling errors and odd greetings is largely obsolete. What still holds:
- An unexpected request to authenticate. The single most reliable signal. Navigate to the service yourself instead of following the link.
- A request that bypasses a normal process. Pay this differently, approve this outside the system, keep this quiet.
- Urgency with a consequence attached. Account closure, a missed payment, a deadline today.
- A reply-to that differs from the From address. Visible in message details, and common in conversation hijacking.
- A link whose destination is not what the text says. Hover and read the host, not the path.
What Makes Awareness Training Effective?
Training that teaches a reporting reflex rather than a detection skill. Nobody detects every phish, and a programme built on the assumption that they should will punish people for being human.
What works: short, frequent, role-relevant content; simulations used to measure reporting rather than to catch people out; and a reporting route that takes one click. What does not: annual slide decks, and simulation results used punitively, which reliably teaches people to stay quiet.
What Should Someone Do If They Click?
Speed matters more than anything else here, and blame destroys speed.
- Say so immediately. The window to contain a credential theft is minutes.
- Change the password from a different device, and revoke active sessions rather than only resetting.
- Check for persistence: new mailbox rules, new forwarding addresses, new app authorisations, new MFA devices. Attackers establish these within minutes of access.
- Preserve the message. Do not delete it; the headers are the evidence.
- Report it, on the timetable in the next section.
A blame-free policy is an operational control rather than a kindness. The organisation that punishes the first report gets the second one an hour later.
What Must You Report, and How Quickly?
In India the clocks are specific, and most people miss both.
| Who | Deadline | Source |
|---|---|---|
| Organisations, for listed incident types | Report to CERT-In "within 6 hours of noticing such incidents or being brought to notice about such incidents" | CERT-In directions, 28 April 2022 |
| Individuals who have lost money | Call 1930, then file on the national portal "within 24 hrs using the acknowledgement number received on SMS" | I4C FAQ |
Two things follow. The six hours is a deadline for telling CERT-In something, not for completing an investigation; an incomplete first report is acceptable and the detail can follow. And the call to 1930 is not the complaint, so a victim who calls and stops has a reference number and no case.
The same CERT-In directions require entities to "enable logs of all their ICT systems and maintain them securely for a rolling period of 180 days", which is what makes a phishing investigation possible at all. Our guides to filing a cyber crime complaint in India and digital forensics investigation cover both sides of that.
What Is the Priority Order?
For a small organisation with limited time, in this order:
- Check your own SPF, DKIM and DMARC with the three commands above. This costs an afternoon.
- Publish DMARC at
p=nonewith a reporting address, read the reports, fix your legitimate senders, then move toquarantineand thenreject. Do not stop at none. - Turn on MFA everywhere, and prefer phishing-resistant factors for administrators and finance.
- Write the payment verification rule down and make it apply to everyone, including whoever is most senior.
- Give people a one-click report button and never punish a report.
- Confirm you have 180 days of logs, because without them an investigation is guesswork.
The first two cost nothing but attention and remove the easiest way to impersonate you. Our guide to card fraud covers what happens after a successful payment phish.
Talk to a WebAsha training advisor about batches, syllabus and current fees.
References
- RFC 7208, Sender Policy Framework (SPF), Version 1, April 2014
- RFC 6376, DomainKeys Identified Mail (DKIM) Signatures, September 2011
- RFC 7489, Domain-based Message Authentication, Reporting, and Conformance (DMARC), March 2015
- CERT-In directions under section 70B(6) of the IT Act, 28 April 2022
- Indian Cyber Crime Coordination Centre, answers to common questions
- National Cyber Crime Reporting Portal
Related reading
Frequently Asked Questions
What's Your Reaction?
Like
0
Dislike
0
Love
0
Funny
0
Wow
0
Sad
0
Angry
0