In brief: Signing in with a code sent by text message is no longer a safe way to protect business accounts. On 10 November 2026 the Centre for Cybersecurity Belgium (CCB) warned that multiple Belgian organisations had been compromised through Microsoft sign-in with SMS as the first factor: a phone number and a one-time code, without a password. SMS and voice codes can be phished, intercepted or redirected through SIM swapping, whether they are the first or the second step. The alternative is phishing-resistant multi-factor authentication (MFA) such as passkeys and FIDO2 security keys. Switching takes preparation, and it only works if employees and the help desk stop handing over codes and resets to whoever asks.

What the Belgian warning says

The CCB advisory is short and direct. A trusted partner observed successful attacks against multiple Belgian organisations in which attackers abused Microsoft authentication with SMS as the first factor. The advisory gives no number of victims, names no organisations and does not attribute the attacks to a group.

The CCB classifies the problem as a misconfiguration, not a software flaw. It concerns SMS and voice authentication in Microsoft Entra ID (formerly Azure AD) for sign-in, MFA and self-service password reset, and any other identity platform or portal that allows SMS or voice as a first factor. Its message: disable SMS as a first factor now, and plan the phase-out of SMS and voice everywhere else.

There is a deadline behind it. Microsoft has announced that it will retire Microsoft-provided SMS and voice authentication in Entra ID from 1 February 2027 for most users, and from 1 July 2027 for Global Administrators and external users.

SMS as a first factor versus SMS as a second factor

The difference matters. With SMS as a second factor, an employee first enters a password and then a code from a text message. This is the classic two-step verification. An attacker needs both.

With SMS as a first factor, the employee types a phone number and receives a code. There is no password. A phone number is not a secret: it is in email signatures and on business cards. Whoever gets hold of one code is in.

That does not make SMS as a second factor safe. In the words of the CCB: “SMS and voice can be phished, intercepted or SIM-swapped, whether used as a first or a second factor.”

Three ways an SMS code fails

Phishing in real time

The employee receives an email or a text message with a link to a copy of the sign-in page. Everything typed there is passed on to the real page within seconds, including the code. A phone call works too: someone poses as IT support and asks the employee to read out the code that “just arrived”. Text messages that lure people to such a page are called smishing, one of the common types of phishing attacks.

Interception

Text messages are not encrypted. The US agency CISA lists the exploitation of weaknesses in SS7, the signalling protocol between telecom networks, as a way to obtain codes sent by SMS or voice. This route is used in targeted attacks.

SIM swapping

The attacker convinces a mobile provider to move the victim’s number to a SIM card he controls. From then on every code arrives on his phone. In 2023 the US Cyber Safety Review Board reviewed the attacks by the Lapsus$ group on large technology companies. It described how the group used fraudulent SIM swaps to take over accounts through sign-in and recovery flows that relied on SMS and voice calls, and it advised organisations to move away from those methods.

The human factor: the code is handed over, not cracked

In two of the three routes no technology is broken. A code is typed into a fake page or read out by an employee who believes he is helping. A SIM swap is social engineering too, only the target is a help desk agent at a telecom provider who wants to assist a customer who “lost his phone”.

Employees have learned that a code on their phone means security. Few have learned that the code itself is the thing an attacker wants, and that nobody from IT, the bank or a supplier ever needs it.

Would your colleagues read a sign-in code out loud to a friendly caller from "IT"?

See how short weekly challenges teach employees that a code is never shared, whoever asks for it.

Discover security awareness training
Free trial available: try Guardey's security awareness training free for 14 days

Which sign-in methods hold up

Not all MFA is equally strong. The comparison below follows the ranking in CISA’s fact sheet Implementing Phishing-Resistant MFA. The US standards body NIST draws the same line in SP 800-63B: a method where a person types over a code is not phishing-resistant.

Sign-in method Resistant to phishing? Resistant to SIM swapping? Remark
SMS or voice code No No Can also be intercepted in the telecom network. CISA: last resort only
Code from an authenticator app No Yes Not tied to a phone number, but the code can still be typed into a fake page or read out
Push notification with number matching No Yes Stops blind approval of repeated prompts (push bombing), not a real-time phishing page
Passkey or FIDO2 security key Yes Yes Only works on the genuine website, so there is no code to hand over

A passkey is a cryptographic key on a phone, a laptop or a physical security key, bound to the address of the real sign-in page. On a fake page it does nothing. CISA calls app codes and push with number matching the best interim option for organisations that cannot switch straight away.

A hand holding a small security key next to an open padlock (illustration)

Moving away from SMS: a step plan

The CCB warns against switching off SMS and voice blindly. Users can be locked out, or left with password-only sign-in. The steps below follow its recommendations. Entra ID is the example, the order applies to any identity platform.

  1. Take stock first. Find out which users, groups, applications and processes rely on SMS or voice: sign-in, MFA, self-service password reset, break-glass accounts, frontline and shared-device workers, external users and third-party systems. In Entra, the SMS and voice usage analyzer and the Authentication methods activity report show this.
  2. Deal with SMS as a first factor with priority. Give those users an alternative and disable it. In Entra: Authentication methods, SMS, turn off “Use for sign-in”. Until then, restrict it to the smallest possible group and monitor that group.
  3. Clean up privileged accounts. Remove SMS and voice from administrator accounts, but only after confirming that a phishing-resistant method works for each of them.
  4. Roll out and test the alternative. Start passkeys or FIDO2 keys with a pilot group. Validate sign-in, password reset, account recovery and the help desk procedures.
  5. Communicate, then phase out. Tell people what changes, when and why, and only then switch off SMS and voice group by group. Do not forget suppliers with access to your systems.
  6. Look back in the logs. Review sign-in and audit logs as far back as retention allows. Look for successful SMS sign-ins, changed phone numbers or methods, unusual locations or devices, and follow-on activity such as new mailbox rules or app consents.
  7. Report what you find. The CCB asks Belgian organisations to report incidents. Entities under NIS2 have to keep the notification deadlines in mind.

What NIS2 says about authentication and training

The NIS2 Directive names authentication in its list of minimum measures. Article 21(2)(j) requires “the use of multi-factor authentication or continuous authentication solutions, secured voice, video and text communications and secured emergency communication systems within the entity, where appropriate”. The words “where appropriate” leave room for a risk-based choice. NIS2 does not literally prescribe phishing-resistant MFA, and it does not ban SMS. After a warning like this one, it is hard to defend SMS sign-in on administrator or internet-facing accounts.

Three other articles are relevant. Article 21(2)(g) requires basic cyber hygiene practices and cybersecurity training. Article 20(2) obliges members of the management body to follow training, which is where NIS2 training for executives comes in. Article 23 sets the deadlines for a significant incident: an early warning within 24 hours, a notification within 72 hours and a final report within one month. Our NIS2 guide explains who is in scope.

Why technology alone does not close the gap

A passkey cannot be phished. The process around it can. Once the front door holds, attackers go for the fallback: a call to the help desk about a lost phone, a recovery method that still runs on a text message, or the registration of a new sign-in method on an account that has already been taken over.

The migration therefore has a human side.

  • Employees need to know that codes and approval prompts are never shared, and what smishing and a real-time phishing page look like.
  • Help desk staff need a fixed way to verify identity before they reset a method, and the backing to say no to an insistent caller.

Phishing simulations, including text message scenarios, let people practise the refusal safely. No law makes them mandatory. They are a practical way to see whether training sticks, and part of what makes an organisation cyber resilient.

Where training fits

Stronger sign-in takes away the code. Training takes away the reflex to help whoever asks for access. You need both.

Guardey offers security awareness training in short weekly challenges, with phishing simulations and reporting per employee on participation and progress. That gives employees regular practice in recognising requests for codes and resets.

Frequently asked questions about SMS authentication

Is SMS authentication still safe?

It is better than a password alone, but it is the weakest form of MFA. SMS codes can be phished, intercepted or redirected through SIM swapping. CISA advises using SMS only as a last resort, and the Belgian CCB advises disabling SMS as a first factor immediately.

What is SIM swapping?

SIM swapping is a form of social engineering in which a criminal convinces a mobile provider to transfer someone’s phone number to a SIM card he controls. Calls and text messages, including sign-in codes, then arrive on the criminal’s phone.

What is phishing-resistant MFA?

Phishing-resistant MFA is a sign-in method in which there is no code or approval that a user can pass on to a fake website. Passkeys and FIDO2 security keys work this way: the key is cryptographically bound to the real sign-in page and does not respond to a copy.

When does Microsoft stop SMS codes in Entra ID?

Microsoft retires Microsoft-provided SMS and voice authentication in Entra ID from 1 February 2027 for most users, and from 1 July 2027 for Global Administrators and external users. Users who only have SMS or voice then have to register a passkey to keep signing in, unless the organisation arranges its own telephony provider.

How would your team react to a text message asking for a sign-in code?

In a personal demo we show how weekly challenges and phishing simulations teach employees never to share a code, and what the reports per employee look like.

Request a demo