Maaz Zindani

Loading the index…

↑↓ move↵ openesc close4072 pages

Help for owners

A staff member keeps getting login approval prompts they did not start

Short answer

Someone probably has the password and is hoping a tired person taps Approve. Tell the person never to approve a prompt they did not start, then change that account's password and sign it out everywhere. CISA recommends phishing-resistant sign-in, or number matching as an interim step for push approvals.

  • Written for business owners
  • Updated
  • By Maaz Zindani, Senior Software Engineer

Does this sound like your situation?

  • A phone keeps showing sign-in approval requests nobody started.
  • The prompts arrive in bursts, often late at night or early morning.
  • Someone then phones or messages claiming to be from IT and asks them to approve one.
  • A staff member approved one just to make the buzzing stop.
  • There are sign-ins from unfamiliar places in the account history.

Why does this happen?

  1. The attacker already has the password

    A prompt only appears after a correct password has been typed. Repeated prompts mean someone is trying to sign in with the real one, taken from a leak, a phishing page or a reused password.

  2. A tap on Approve is easy to give away

    CISA says organisations that use push-notification sign-in and cannot yet adopt phishing-resistant methods should use number matching to mitigate this tactic, called MFA fatigue.

  3. Not all second steps are equally strong

    CISA ranks a security key as the strongest, then an authenticator app with number matching, then app codes, biometrics, and puts text or email codes last, to be used only when stronger options are not available.

  4. Number matching is described as a stop-gap

    CISA calls number matching a more secure option than mobile push, for an interim step while planning your move to phishing-resistant MFA. It is not as strong as phishing-resistant methods.

  5. The password and the account are still exposed

    While the password is known, the attacker can keep trying and may also be trying other accounts that share it. The NCSC advises changing the password on the account and on any account using the same one.

What can I do about it today?

  1. Tell the person to deny every promptSay clearly: if you did not start it, tap Deny or ignore it, whoever phones or messages you afterwards. A genuine colleague has no reason to ask you to approve a prompt you did not start.
  2. Change the account password straight awayUse a new password that is not used anywhere else, typed on a device you trust.
  3. Sign out all devices and sessionsUse the service's setting that ends other sessions, so a sign-in that already succeeded is cut off.
  4. Look at the recent sign-in historyCheck for sign-ins from places or devices the person does not recognise. If you see an approved one, treat the account as taken over.
  5. Check whether the person approved oneAsk gently, without blame. If yes, also check forwarding rules and connected apps in the mailbox.
  6. Warn the rest of the teamA short message: if prompts arrive that you did not start, deny them and tell me. Do not name the person.

When should I bring someone in?

  • A prompt was approved, or you cannot tell.
  • The account has admin rights, payment access or access to servers.
  • Several staff are receiving prompts at the same time.

How do I handle it?

  1. Contain the accountPassword reset, session end and review of sign-in logs, recent changes, forwarding rules and connected apps.
  2. Decide where the password leaked fromReuse, a phishing page or a breach on another service, so the same password is retired everywhere.
  3. Switch to a sign-in step that is harder to trickWhere the service supports it, security keys or passkeys, otherwise number matching, with the weakest options switched off.
  4. Limit the prompts themselvesWhere the service allows, I set a limit on repeated requests and turn on alerts for sign-in attempts from new places.
  5. Brief staff in one pageWhat a legitimate prompt looks like, what to do with one that is not, and who to tell.

The price depends on what I find, so I quote in writing after a short discovery call. Tell me what is happening and I reply within 24 hours.

What do owners ask about this?

Is my two-step verification broken?

No. It is doing its job, because the prompt stops the attacker unless someone approves. The risk is a person tapping Approve, which is why stronger methods exist.

Is number matching enough?

CISA treats it as an interim step. Phishing-resistant methods such as security keys are stronger.

Should I switch the second step off to stop the noise?

No. That would remove the barrier in front of the account. Change the password and end sessions instead.

Who is behind it?

I cannot tell from the prompts alone. The sign-in log can show where attempts came from, but the useful action is the same either way.

Sources

How was this page made?

I drafted this page with AI assistance from the official documents listed under Sources. Found a mistake? Tell me.

Next step

Want this off your desk?

Send me a few lines on where things stand. I reply within 24 hours.