Topic · Threats
Phishing
Phishing is a deceptive message or interaction intended to obtain credentials, money or sensitive information.
How it works
Attackers impersonate trusted senders or create urgency through links, attachments and conversation tactics.
What can go wrong
Authentication can stop some domain spoofing but cannot detect every compromised or lookalike sender.
What to check or do
Train reporting workflows, protect accounts with MFA and validate unusual requests.
Practical scenario
A recipient receives a realistic invoice request linking to a site that visually copies a supplier’s portal. The sender display name resembles the supplier, but the actual domain differs. A technically authenticated attacker-controlled domain is not proof the invoice is genuine.
Verification checklist
- Inspect full sender and destination domains without trusting display names.
- Verify urgent payment or credential requests through a known independent channel.
- Preserve full headers, URLs and screenshots for investigation.
- Use user reporting and domain monitoring alongside SPF, DKIM and DMARC controls.
How to investigate Phishing
- Signal to investigate
- A convincing login page appears in an email that passes authentication for an attacker-owned domain.
- Evidence to collect
- Inspect requested action, destination URL, domain ownership and account-sign-in evidence.
- Common wrong conclusion
- Passing SPF, DKIM and DMARC does not make an unrelated malicious sender safe.
How to interpret the evidence
A malicious actor may send authenticated messages from a domain they control, or use a real account compromised through stolen credentials or tokens. Consequently a passing authentication summary cannot distinguish benign from harmful intent. The meaningful signal may be an anomalous request, deceptive URL, suspicious attachment or changed payment instruction.
What to change and how to verify it
Verify sensitive actions through independent channels. Preserve the original message and evidence, isolate unsafe URLs, and review affected account/session activity when compromise is plausible. Report or block domains using accurate artifacts instead of relying on visual brand similarity alone.
Put this into practice
Run a focused check, then explore protection if your domain needs it.