bastion

Common Authentication and Access Threats

Build this skill

Recognise the signs of credential theft, account takeover, session hijacking and account recovery attacks, and choose the defence that blocks each one.

The night someone else became Mrs. Gokhale

Mrs. Lata Gokhale opens the Tulsi Health app at 8 in the morning to reorder her blood pressure tablets and finds things she never did. Her mobile number was changed at 1:47 am, her delivery address moved from Kothrud to Wakad, and an order worth ₹18,400 was placed with her saved card. Her phone also shows "No service". Nobody broke the app's code. Someone simply logged in as her. Most attacks on accounts work like this: rather than smashing through a wall, the attacker gets hold of the thing that proves who you are and walks in through the front door. Knowing the handful of ways that happens is how you notice it early and block it.

Mrs. Lata Gokhale checks her app at 8 am 8:02 No service Tulsi Health app Account activity Mobile number changed 1:47 am · new device Address: Wakad, Pune 1:52 am · was Kothrud Order ₹18,400 placed 2:10 am · saved card Monthly BP medicines 12 Oct · her usual order ! Her SIM stopped working Someone may have moved her number to a new SIM to receive her OTPs ! Contact details changed first so alerts now go to the attacker ! Odd hour, new address, big order Medicines to be resold, paid with her saved card ✓ Her normal pattern Daytime, home address, ₹1,200

An account takeover leaves a trail: contact details changed first, then the account used in ways that do not match its owner's normal habits.

Credential theft: stealing the proof of who you are

A credential is whatever you present to prove you are you: a username and password, a one-time code (OTP) sent by SMS, a fingerprint, or a passkey stored on your phone. Credential theft is getting hold of someone else's credentials. It rarely needs clever technology. Phishing pages copy a real login screen and collect what people type into them. Malware (harmful software) on a laptop can record keystrokes. Passwords are read off sticky notes or over a shoulder at a busy counter. And when any website where you used a password is breached, attackers try that same email and password on hundreds of other sites, a technique called credential stuffing. It works because so many people reuse one password everywhere.

The analogy is a house key. A burglar does not need to pick the lock if your key is under the doormat, if you lent a copy to someone careless, or if the same key opens your house, your office and your scooter. Mrs. Gokhale had signed up on a recipe website that was breached last year, so her name, date of birth, address and old password were already circulating among criminals, and stolen details like these are the raw material for every attack that follows.

Account takeover and unauthorized access

Account takeover is what happens next: an attacker uses stolen credentials to control a real person's account. Because the logins are genuine, the system sees nothing wrong at first. Attackers usually act in a set order. They change the contact details so that warnings go to them, then they use whatever the account is worth: saved cards, loyalty points, personal data, or the trust of colleagues who will believe an email from that address.

Unauthorized access is the broader idea: anyone reaching data or systems they are not permitted to reach, whether by a takeover or not. An ex-employee whose account was never closed, a delivery boy who uses a store laptop left logged in, and an employee who looks up a neighbour's prescriptions out of curiosity are all unauthorized access. In each case the owner of the data did not agree to it, and that is what makes it a security incident.

Access threatsgetting in as someoneelsesteal proofCredential theftpassword, OTP stolenskip loginSession hijackbadge after logingo aroundRecovery attackfake a resetleft openUnused accountsold or shared logins

Every route ends the same way, with someone acting as a person they are not. Strong passwords guard only the first route.

Session hijacking: the stolen visitor badge

When you visit Tulsi Health's head office, the receptionist checks your ID card once and hands you a visitor badge. For the rest of the day, guards look only at the badge, not at your ID again. Websites and apps work the same way. After you log in, the server (the computer that runs the website) gives your browser a session token, a long random value usually kept in a cookie, which is a small piece of data the website stores in your browser. Every later click carries this token, so the site knows it is still you without asking for the password each time. The period during which the token stays valid is called a session.

Lata's browserTulsi Health appPassword + OTP1Session token (the badge)2View orders + token3Reorder tablets + token4

After step 2, the token alone proves who Lata is. Whoever holds a copy of it is treated as Lata until it expires.

Session hijacking is stealing or copying that token and using it to act as the logged-in person. The thief never learns the password and never sees the OTP, which is exactly why it is dangerous: the login protections have already been passed. Tokens are typically exposed through malware on the device, through a fake or unprotected website connection, or through a shared computer where the previous user never logged out. Defences belong mostly to the website: sending everything over encrypted connections (the padlock, HTTPS), making tokens expire after a period of inactivity, cancelling all tokens when the password changes, and asking for the password again before risky actions such as changing a mobile number. Your part is simple: log out on shared computers and keep your devices free of malware.

Account recovery attacks: going around the front door

Every account has a "Forgot password?" path, because real people do forget. An account recovery attack abuses that path instead of the login. Picture someone walking into your bank branch, claiming to be you, saying they lost their passbook, and talking a busy clerk into issuing a new one. The vault was never touched; the attacker went through the help desk.

Online, the same trick takes a few forms. Security questions such as "your mother's maiden name" or "your first school" can often be answered from social media. A support agent can be pressured with a sad story. The most serious form is a SIM swap: the attacker convinces a mobile operator's store or helpline that they are you and that your SIM was lost, using stolen or forged identity details. The operator moves your number to a new SIM in the attacker's hand, your own phone shows "No service", and every OTP for "forgot password" now arrives with the attacker. That is what happened to Mrs. Gokhale at 1:47 am.

Defences make recovery as strong as the login itself: an authenticator app or passkey rather than SMS alone, alerts sent to the old contact details whenever they change, a waiting period before a new number can be used for big actions, and staff trained to verify callers through a known number. If your phone suddenly loses signal for no reason, call your operator from another phone straight away. Step through Mrs. Gokhale's night below and predict each move before it is revealed.

Reading the signs in a login log

Most takeovers are caught by someone reading a log: a record the system keeps of every login and important change, with the time, the account and where it came from. One suspicious line rarely proves anything. A pattern does: a burst of failed logins, a success from an unfamiliar place or device, contact details changed soon after, then unusual activity at an odd hour. Neha reviews the app's log for the night below. Tick the lines that point to an account takeover and leave the normal ones alone.

Follow your curiosity.

↑ ↓ to explore · Enter to open · Escape to close