Social Engineering and Phishing: How to Spot and Stop Them
The most reliable exploit in cyber security is not a software bug; it is human nature. Social engineering manipulates people into revealing information or taking actions that help an attacker, and phishing is its most common form. Technical controls help, but people remain the last line of defence, which means every security professional needs to understand how these attacks work and how to build resilience against them.
Why It Works
Attackers exploit predictable psychological triggers:
- Authority: the message appears to come from a boss, bank or government agency.
- Urgency and fear: “Your account will be closed in 24 hours.”
- Greed and curiosity: prizes, refunds, scandalous attachments.
- Helpfulness and trust: people want to assist a colleague or customer.
- Social proof: “Everyone in your team has already completed this form.”
- Familiarity: mimicking a known brand, tone or contact.
Recognising these levers is half of the defence. Anyone, including security experts, can be fooled when busy, tired or distracted. The goal is to build processes that do not depend on perfect attention.
Types of Social Engineering
| Type | Channel | Example |
|---|---|---|
| Phishing | A fake bank message with a link to a counterfeit login page | |
| Spear phishing | Email, targeted | A message to a finance employee that references a real project and manager |
| Whaling | Email, executives | A fake legal notice sent to a CEO |
| Business email compromise (BEC) | An “urgent” instruction from a spoofed or compromised executive to change bank details or pay an invoice | |
| Smishing | SMS | A fake courier or bank text with a link |
| Vishing | Phone | A caller posing as bank or IT support asking for a code or remote access |
| Pretexting | Any | An invented scenario, such as a “new auditor” needing documents |
| Quishing | QR codes | A malicious QR code on a poster or in an email |
| Baiting | Physical or digital | A USB drive left in a car park labelled “Salaries” |
| Tailgating | Physical | Following an employee through a secure door |
| Impersonation of support | Phone, chat | Asking a help desk to reset someone else’s password |
In India, common consumer scams include fake KYC-update messages, fake courier and customs fee notices, “UPI collect request” tricks (where a victim is asked to approve a payment request while believing money is coming in), fake job offers requiring a registration fee, and impersonation of bank or government officials. The principles below defend against all of them.
Anatomy of a Phishing Email
Here is a fictional example to analyse. The domains are invented for teaching purposes.
From: "HR Payroll Team" <payroll-update@secure-hr-portal.example.net>
To: all-staff@yourcompany.example
Subject: URGENT: Salary revision - action required today
Dear Employee,
Your salary revision letter is ready. To avoid delay in your October
payment, confirm your details within 2 hours using the secure link below.
https://yourcompany-hr-login.example.net/confirm
Failure to comply will result in payment being withheld.
HR Department
Red flags a trained reader should spot:
- Mismatched sender domain: the company’s real domain is
yourcompany.example, but the message comes fromsecure-hr-portal.example.net. - Urgency and threat: a two-hour deadline and withheld payment.
- Generic greeting instead of a name.
- Look-alike link: the domain is not the company’s; hovering over a link reveals its true destination.
- Request for credentials through a link instead of via the known internal portal.
- Unusual process for a payroll matter.
Email Authentication: SPF, DKIM and DMARC
Three DNS-based standards help receiving mail servers decide whether a message really came from the domain it claims:
- SPF: lists which servers may send mail for a domain.
- DKIM: adds a cryptographic signature to messages, verified using a public key in DNS.
- DMARC: tells receivers what to do when SPF and DKIM checks fail to align with the visible From domain (none, quarantine or reject) and sends reports to the domain owner.
Look at the headers of a message (in most mail clients, choose “Show original” or “View source”). A line similar to the following summarises the checks:
Authentication-Results: mx.example.org;
spf=fail smtp.mailfrom=secure-hr-portal.example.net;
dkim=none;
dmarc=fail header.from=secure-hr-portal.example.net
Failing results do not prove malice and passing results do not prove safety (attackers can register their own look-alike domains and pass all checks for them), but they are valuable signals. For your own domain, publish SPF, DKIM and a DMARC policy, moving gradually toward p=reject to stop others spoofing your name. You can check your published records:
dig +short TXT yourdomain.example
dig +short TXT _dmarc.yourdomain.example
Safe Practice: Analysing a Suspicious Message
- Do not click or open attachments.
- Check the sender’s real address, not just the display name.
- Hover over links (without clicking) and compare the domain with the real one. Look for subtle changes such as
rninstead ofm. - Check the headers for authentication results and the originating path.
- Look for emotional pressure and unusual requests.
- Verify through a separate, known channel: call the real number from the company website, not the one in the message.
- Report it through your organisation’s phishing-report button or security team, then delete it.
Reputation services such as VirusTotal and URLScan can check a suspicious URL, but never submit confidential or personal content to public services. Open suspicious links only in an isolated lab VM and never enter real credentials.
Defences That Work
Technical controls
- Email filtering with attachment sandboxing and link rewriting.
- SPF, DKIM and DMARC on your own domains; warning banners for external senders.
- Phishing-resistant MFA (security keys, passkeys), which stops stolen passwords from working on fake sites; see Password Security.
- DNS and web filtering to block known malicious domains.
- Endpoint protection and macro and script restrictions.
- Browser and password-manager autofill, which refuses to fill credentials on look-alike domains.
Process controls
- Out-of-band verification for any change of bank details or urgent payment request: call back on a known number.
- Dual approval for payments above a threshold.
- Strict identity verification before help-desk password resets.
- Clear, easy reporting channels, with thanks and not blame for reports.
People and culture
- Short, regular, relevant training rather than annual lectures.
- Phishing simulations used to teach, not to punish. Measure the reporting rate as well as the click rate.
- Leaders modelling good behaviour, including reporting their own mistakes quickly.
If You Have Clicked or Shared Something
- Disconnect from the network if you opened an attachment, and tell your security team immediately.
- Change the password from a different, clean device and revoke active sessions.
- If you entered a card or banking detail, contact the bank at once using the number on your card or the official app.
- In India, report financial fraud promptly at the national cybercrime helpline 1930 or at cybercrime.gov.in; fast reporting improves the chance of stopping a transfer.
- Watch accounts for unusual activity and enable MFA everywhere.
Practice Exercises
- Write down three real phishing or scam messages you or friends have received (remove personal details). Identify the psychological trigger and the red flags in each.
- Create a one-page awareness poster for your family or team showing how to verify an “urgent” request.
- Check your own domain or an email provider’s documentation to find out how SPF, DKIM and DMARC work and how to read headers.
- Design a callback procedure for a finance team that receives a request to change a supplier’s bank account.
- Explain why a padlock in the address bar does not make a website trustworthy.
Mentor tip: Never shame the person who clicked. The most secure organisations are the ones where people report mistakes within minutes, because speed of reporting often decides the size of the damage.
Educational use only: run phishing simulations only with organisational authorisation, and never target real people without consent.
- Social engineering exploits trust, urgency, authority and curiosity rather than software flaws.
- Learn the red flags: mismatched domains, pressure, unusual requests and look-alike links.
- SPF, DKIM and DMARC protect your domain; phishing-resistant MFA protects accounts.
- Out-of-band verification and dual approval stop most payment fraud.
- Make reporting easy and blameless; speed of reporting reduces damage.