In this article
Introduction
In 2024, a state-backed hacking group breached Microsoft's own corporate network. Not through some novel zero-day exploit, but by finding a single test account that didn't have multi-factor authentication enabled, guessing its password through simple password spraying, and using that foothold to read executive email for over a month before anyone noticed. If the company that literally builds enterprise identity security can be caught out by one account without MFA, the case for treating it as optional anywhere in an organization is hard to make. This guide covers what MFA actually is, what the evidence says about its effectiveness, why not all MFA methods are equally trustworthy, and how to roll it out properly.
What MFA Actually Is
Multi-factor authentication requires a user to prove their identity using two or more factors drawn from different categories, rather than a password alone.
The three factor types
Knowledge
Something you know
A password, PIN, or answer to a security question.
Possession
Something you have
A phone receiving a code, an authenticator app, or a physical hardware security key.
Inherence
Something you are
A fingerprint, facial scan, or other biometric.
Genuine MFA combines factors from at least two of these categories. A password plus a security question isn't MFA, both are knowledge factors. A password plus a code from an authenticator app is, because it combines knowledge and possession.
The Evidence: Why It's Not Optional Anymore
The case for MFA, by the numbers
of account compromise attacks blocked by MFA, according to Microsoft's own research
of accounts Microsoft observed being compromised had no MFA enabled
password attacks per second Microsoft blocks across Entra ID
increase in identity-based attacks in the first half of 2025 alone (Microsoft Digital Defense Report 2025)
The scale of the attack volume is what makes the effectiveness number matter so much. Microsoft has stated it processes over 600 million identity attacks per day. Against that volume, a single control blocking more than 99% of compromise attempts isn't a marginal improvement, it's close to the difference between an account being a realistic target and not being one.
Real-World Proof: When Microsoft Itself Got Breached
Statistics about attack-blocking rates can feel abstract until there's a concrete case attached. In 2024, the Russian state-backed group known as Midnight Blizzard compromised Microsoft's own corporate network by targeting a legacy, non-production test account that had never been enrolled in MFA. Using simple password spraying, a low-sophistication technique that simply tries common or previously breached passwords against many accounts, the group gained a foothold. From there, they used that access to compromise a small number of Microsoft corporate email accounts, including members of the senior leadership team, and read email for more than a month before the intrusion was detected. Nothing about the attack itself was technically sophisticated. It succeeded because of one missing control on one account.
Not All MFA Is Equal: The Strength Hierarchy
MFA strength hierarchy, weakest to strongest
SMS / Email codes
- Security level
- Weakest
- Vulnerable to
- Interception, SIM swapping, phishing
- Why
- Codes can be intercepted or relayed
- Regulatory trend
- Being phased out for banking by several financial regulators
Authenticator app / Push
- Security level
- Moderate
- Vulnerable to
- Real-time phishing, push and prompt bombing
- Why
- Still relies on a code or approval a user can be tricked into providing
- Regulatory trend
- Widely accepted, increasingly considered a floor rather than a ceiling
Phishing-resistant (FIDO2, passkeys)
- Security level
- Strongest
- Vulnerable to
- Effectively immune to remote phishing
- Why
- Cryptographically bound to the legitimate site and device; there's no code to steal
- Regulatory trend
- Actively promoted by CISA and major platforms as the target standard
The takeaway isn't that SMS-based MFA is worthless. Population-level data consistently shows even basic MFA blocks the vast majority of automated, password-based attacks. It's that as attackers get more sophisticated, risk increasingly concentrates in the weaker methods, and phishing-resistant MFA is what closes that remaining gap.
When Even MFA Gets Bypassed
MFA is not a silver bullet, and understanding how it fails matters as much as knowing that it works. Two attack patterns specifically target weaker MFA implementations. MFA fatigue, or push bombing, is when an attacker who already has a valid password repeatedly triggers push notifications until a user, out of confusion, annoyance, or assuming it's a glitch, approves one by mistake. This works against push-based MFA but not against phishing-resistant hardware keys, since there's no prompt to spam. Adversary-in-the-middle phishing is when a fake login page relays a victim's password and one-time code to the real site in real time, capturing a valid session. This defeats SMS and authenticator app codes but not phishing-resistant MFA, which cryptographically verifies it's talking to the legitimate site before completing authentication.
Where Mandates Are Already Here
This has moved well past best-practice recommendation into active enforcement by major platforms.
MFA mandates already in motion
- 2024
Microsoft begins enforcing mandatory MFA for Azure sign-ins.
- February 2026
Microsoft enforces mandatory MFA for all Microsoft 365 admin center sign-ins, with no exceptions or grace period after the deadline.
- July 2026
Deadline for Microsoft's Phase 2 mandatory MFA enforcement across tenants.
- Ongoing
Several financial regulators have moved to phase out SMS-based one-time codes for banking specifically, pushing the industry toward stronger methods.
Major providers including Salesforce, GitHub, AWS, and Microsoft have all committed to mandatory MFA for privileged accounts, with phishing-resistant methods expected to follow as the next baseline.
The Adoption Gap: Why Isn't Everyone Doing This Already?
Given evidence this consistent, the remaining adoption gap is worth naming honestly. Okta's Secure Sign-in Trends Report 2025 found workforce MFA adoption at roughly 70% as of January 2025, meaningful progress, but still leaving a substantial share of accounts unprotected by a control that blocks the overwhelming majority of automated attacks. The gap tends to come down to friction: extra steps at login, unfamiliar apps, and lost-device recovery concerns all create real, if manageable, resistance. Newer methods like passkeys are specifically designed to reduce that friction while keeping the security benefit, which is part of why the FIDO Alliance's State of Passkeys 2026 report found 68% of surveyed organizations already deploying or actively rolling out passkeys for workforce sign-in.
A Practical MFA Rollout Framework
MFA rollout checklist
- 01Start with privileged and admin accounts. These carry the highest risk and should be the first, non-negotiable priority, not the last.
- 02Move away from SMS where possible. Use it as a fallback rather than the primary method, especially for anything handling sensitive data or financial transactions.
- 03Prioritize phishing-resistant methods for high-risk roles. Finance, IT administration, and executive accounts are worth the modest additional friction of hardware keys or passkeys.
- 04Plan for lost-device and recovery scenarios up front. Poor recovery processes are one of the most common reasons users resist MFA or find workarounds that weaken it.
- 05Treat MFA rollout as change management, not just IT configuration. Clear communication about why it matters reduces resistance more effectively than a mandate alone.
MFA is one of the five pillars covered in our guide to identity and access management, and it sits alongside the longer-horizon work of post-quantum cryptography in any serious security roadmap.
Sources
- Microsoft, "Plan for mandatory Microsoft Entra multifactor authentication," Microsoft Learn (learn.microsoft.com)
- Microsoft, Digital Defense Report 2025
- Okta, Secure Sign-in Trends Report 2025
- FIDO Alliance, State of Passkeys 2026 report
This article reflects publicly available research and vendor announcements as of July 2026. MFA enforcement deadlines and adoption figures are updated regularly by these organizations, verify current details before finalizing a rollout plan.
Frequently Asked Questions
What is multi-factor authentication?+
MFA requires a user to provide two or more independent forms of proof before accessing an account: something they know, something they have, or something they are.
How effective is MFA at stopping attacks?+
Microsoft's research found MFA blocks more than 99.2% of account compromise attacks, and more than 99.9% of accounts Microsoft observed being compromised had no MFA enabled at all.
What are the three types of authentication factors?+
Knowledge (something you know), possession (something you have), and inherence (something you are). MFA requires combining factors from at least two different categories.
Is all MFA equally secure?+
No. SMS and email codes are weakest, authenticator app codes and push notifications are stronger but still phishable, and phishing-resistant methods like FIDO2 security keys and passkeys are strongest.
What is an MFA fatigue attack?+
Also called push bombing, it's when an attacker with a valid password repeatedly triggers MFA push notifications until a user approves one by mistake. It works against push-based MFA but not phishing-resistant methods.
Should businesses use SMS-based MFA?+
It's far better than no MFA, but it's the weakest widely available option, and several financial regulators are phasing it out for banking. Prioritize stronger methods where possible.
