Microsoft's November 10 Patch Tuesday fixed CVE-2026-124194, an Outlook spoofing bug with a CVSS score of 6.5. In a typical patch review, a 6.5 lands in the middle of the spreadsheet, below the criticals and the remote code execution bugs that get the headlines. This one belongs at the top. CSO Online's roundup says threat actors are already attacking it.

Here's the mechanism. The bug hands over the user's NTLMv2 password hash. From there the attacker can relay it to sign in as that user, or take it offline and crack it.

I think the 'spoofing' label does almost as much damage as the score. Spoofing sounds cosmetic, like a faked sender name on an email. This bug gives away a credential, and I'd bet plenty of admins skimmed past it for exactly that reason.


What's on the record, and what isn't

CSO Online's Patch Tuesday roundup (https://www.csoonline.com/article/4247015/november-2026-patch-tuesday-63-critical-microsoft-fixes-plus-an-outlook-zero-day-that-allows-theft-of-hashed-credentials.html) reports that the vulnerability affects Microsoft 365 Apps for Enterprise and Office LTSC 2021 and 2024. Microsoft classifies it as spoofing and rates it 6.5. The same roundup calls it a currently exploited zero-day and says administrators need to address it immediately because threat actors are already attacking it.

That's the extent of the public record I'm working from.


Why a credential leak scores 6.5

This bug needs a user to open a file. My read is that the user-interaction requirement keeps the number in the middle of the range. That's a fair description of the bug.

They're poor descriptions of the risk. Users open files all day. That's their job. An attack that needs one click from someone in accounts payable clears the bar in almost every organization, and a phishing campaign only needs one person out of fifty to open the attachment.


What an NTLMv2 hash gets an attacker

If you run on-premises Active Directory, something in your environment almost certainly still accepts NTLM, and most small shops have never inventoried where.

Cracking is the path that worries me more for smaller companies. A weak or reused password falls fast.


  • Work out which update channel each group of machines sits on.
  • Chase the stragglers. A laptop that sleeps instead of shutting down can sit unpatched for weeks while its owner leaves Outlook open.
  • Hunt for Office installs you don't manage: a contractor's machine, a reception PC someone built years ago, the owner's personal laptop with the company mailbox on it. That last one is often the hardest conversation.
  • Block outbound SMB at the firewall if you haven't already. Almost no small business needs workstations talking SMB to the internet.
  • Confirm MFA covers every account and that Conditional Access blocks legacy authentication. A cracked password with no MFA behind it is an account takeover.
  • Send users a short note: Outlook bug, be careful with attachments, the patch is rolling out, report anything odd. Skip the fear.

SMB and LDAP signing, and an NTLM audit

Enforce SMB signing on servers and turn on LDAP signing and channel binding on domain controllers.

Audit NTLM use. Run it for a few weeks and you'll know which systems would break if you restricted it. Of the two, this is the one I'd start today.


Sort by exploitation first, score second

My patch rule has two inputs. Anything on the KEV catalog, or anything the vendor marks as exploited, goes to the top regardless of score. After that, sort by score and by how exposed the affected product is. Outlook sits on every desk and opens attacker-supplied content all day, so it ranks high on exposure even when the CVSS number doesn't.

Base score still matters for the long tail of bugs nobody is exploiting. Against evidence of real attacks, it's a poor tiebreaker.


A 6.5 with a working exploit in the wild is more urgent than a 9.8 that nobody has weaponized. CVE-2026-124194 needs a click, leaks a hash, and lands in the moderate band, and CSO Online reports attackers are already using it.

Patch, and use the same week to close outbound SMB and check MFA coverage, so the next Outlook hash leak has less to work with.


If you've already pushed this patch: how long did your slowest machines take to pick up the fixed Outlook build, and what held them back? Sleeping laptops, a stuck Semi-Annual Enterprise Channel, or users who never close Outlook? I suspect the last one, and I'd like to know if I'm wrong.