bastion

Multifactor Authentication and Passkeys

Build this skill

Classify login proofs as know, have or are, compare SMS codes, authenticator apps and passkeys, and explain why passkeys resist fake login pages.

The login that stopped at the second step

At 3 am one Sunday, someone in Bhopal typed Sameer Joshi's email address and his correct password into the Tulsi Health mail login. The password had leaked from another website, so as far as the password check was concerned, this was Sameer. But the mail service then asked for a six-digit code from the authenticator app on Sameer's phone, and the person in Bhopal did not have his phone. The login stopped there. Asking for a second, different kind of proof is called multifactor authentication, and it is the single most effective step most people can add to an important account.

Three kinds of proof

Each kind of proof you can offer a system is called an authentication factor, and there are three kinds. Something you know is a secret in your head: a password, a PIN, the answer to a security question. Something you have is a physical object: your phone, a bank card, a small USB security key, an office ID card. Something you are is a feature of your body, a biometric: your fingerprint, your face, your voice. Each kind fails in a different way. A password can be guessed or tricked out of you without anyone touching you, a phone has to be physically taken or hijacked, and a fingerprint is very hard to copy from a distance.

A bank locker shows why mixing kinds works so well. To open one, you need the locker number and a secret word, your own locker key, and the bank's master key turned alongside yours, and the officer compares your signature and face with the register. A thief who steals your key from your cupboard at home still has to look like you and sign like you in front of a bank officer. Each extra check is of a different kind, so one theft is not enough.

Bank locker room No. 214 opens only when every check passes ? Something you know Locker number and the secret word like a password or a PIN Something you have Your locker key, plus the bank's own key like your phone or a security key Something you are Your signature and face, checked against the register like a fingerprint or face scan A thief with your key still needs your signature. Different kinds of proof fail differently.

A bank locker asks for proof of three different kinds. A thief who has one kind still lacks the others, and that is the whole idea of multifactor authentication.

The kinds matter more than the count. A password plus a security question is two things you know, so it is still single-factor: whoever tricked you into revealing one can usually trick you into revealing the other. Sort the proofs below into their kinds to practise telling them apart.

Two-factor and multifactor authentication

Two-factor authentication, or 2FA, means a login needs exactly two different kinds of proof, most often a password plus something you have. Multifactor authentication, or MFA, is the general term for two or more. You meet it every day: an ATM wants your card and your PIN, and a UPI payment needs your phone with its registered SIM and your UPI PIN.

Not all second steps are equally strong. An SMS code is better than nothing, but it can be stolen through a SIM swap, where a criminal convinces a mobile shop to move your number onto a new SIM card, after which your codes arrive on their phone. Push approvals, where an app simply asks "Is this you trying to sign in?", can be abused by sending prompt after prompt at night until a tired person taps Approve; the defence is never to approve a login you did not just start, and to prefer apps that make you type a number shown on the login screen. Codes from an authenticator app avoid both problems.

Passkey or security keyrefuses lookalike sitesstrongestAuthenticator app codeoffline, changes every 30 sPush approvalrisky if you tap without thinkingSMS codeSIM swap can steal itPassword onlyone leak is enoughweakest

Any second step is a big improvement over a password alone. The higher steps also resist SIM swaps, prompt flooding and, at the top, fake login pages.

How an authenticator app makes its codes

When Sameer turned on the second step for his mail, the website showed a QR code, a square pattern of dots that a phone camera can read. His authenticator app scanned it and saved the secret hidden inside, and the website kept its own copy of the same secret. From then on, the app and the website each combine that secret with the current time and work out the same six digits, which change every 30 seconds. The app needs no network and no SMS, which is why it keeps working in a basement storeroom with no signal.

Authenticator Tulsi Health mail sameer.joshi 482 913 College portal sameer.j 075 264 new codes every 30 seconds A code that changes every 30 s Sign in Use your passkey for tulsihealth.in Touch the sensor your fingerprint stays on this phone Nothing to type, nothing to copy Authenticator code phone and site share a secret; the code is made from it and the time ! can be typed into a fake Passkey the phone checks the site's real name, then signs with its own key ✓ refuses a lookalike

Both methods prove you have your phone. The difference is that a code is something you can be tricked into typing, while a passkey is never typed at all.

Codes still have one weakness: a person has to type them somewhere. If a fake login page at tulsihealth-secure.in asks for your password and then your code, the attacker behind it can pass both to the real site within seconds, well inside the 30-second window. Step through the three versions below and predict what happens when the second step is a code, and when it is a passkey.

Passkeys: a key that knows its own door

A passkey replaces the password altogether. When you create one for a website, your phone or laptop makes a matched pair of keys. The private key never leaves your device and is locked behind your fingerprint, face or screen PIN. The public key is given to the website, and it can only check signatures, never make them. Passkeys are built on an open standard called FIDO2, so the same idea works across phones, laptops and websites from different companies.

To log in, the website sends your device a fresh random challenge. You unlock with your fingerprint, and the device signs the challenge with the private key; the website checks the signature with the public key it stored. Your fingerprint is only used to unlock the key on your own device and is never sent to the website. Because the website stores only public keys, a stolen copy of its database gives attackers nothing they can log in with.

SameerPhonetulsihealth.inrandom challenge1touch to sign in?2fingerprint (stays here)3challenge signed4check with public key5

Only the signed challenge travels to the website. The private key and the fingerprint never leave the phone, so there is nothing for a fake page or a database thief to collect.

The protection against fake pages comes from one detail: the device remembers which website each passkey belongs to and will only use it on that exact address. It is like a key cut so that it turns only in your own front door, and simply will not enter a lookalike door painted to match. You cannot be tricked into typing a passkey, because there is nothing to type. If your phone is lost, passkeys are usually backed up through your phone's account and can be restored to a new phone, so keep that account itself protected with a strong passphrase and a second step.

Follow your curiosity.

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