Microsoft got hacked. A signing key was compromised that allowed attackers to create their own access tokens.
How:
Some theories —
Microsoft isn’t saying how the key was obtained
A - vuln in O365 that allowed attackers to get a key
B - compromise of internal microsoft resources (phising attack, etc). compromise of the cloud service
C - used an account to create other certificates (somehow)
D - freckin GitHub
(John) It sounds like, conspiratorially, that we might’ve been in the Chinese networks, through the NSA, NRO, FBI, etc.., then they contacted Microsoft. BILL BILL caught them haha points to state department.
US government detected weird mail access logs…apparently
When:
This wasn’t detected by Microsoft but rather by the state department
June 16th
More:
People probably want to do audits with their logs, so they will have to put something out soon
Microsoft will contact people who may have been affected, but they aren’t saying much
Why do logs cost extra tiers on products lol?
it’s stupid. this is a necessary commodity
we have to push back to get logs and choose platforms with better logging
the problem is that logs require computation and everyone is stingy with cloud resources and making money off it
Any tool that allows to traversal that uses and operates on PII with no authentication is usually a bad idea
Companies should be careful in the same manner on authenticating for PII with their APIs when it involves GET requests to supposedly public information
SIM swapping is a simple attack that involves tricking or bribing companies for SIM swaps
This allows interception of account resets, MFA, and more which can be used to gain access to financial accounts or accounts with data that can be extorted against the victim
The lesson: get burner phones or figure out how to get rid of the weak link of phone numbers
It’s common for websites and local web apps to request access to local resources to fingerprint users or collect info about what software runs on the user’s machine.
Most browsers allows websites local resource access just as easily as other resources with just the browser.
This practice is typically documented on websites:
examples - eBay, Citibank, Chick-fil-A, etc.
usually anti-fraud scripts
Safari blocks these requests even from secure public websites as a side-effect of its security measures, rather than on purpose
Brave is introducing a localhost access permission for granular control of this with websites
“Brave is the only browser that will block requests to localhost resources from both secure and insecure public sites, while still maintaining a compatibility path for sites that users trust,” pledges the Brave team.
Brave will maintain allow-lists for trusted sites that can ask for this permission
Process injection is looked after by OS companies, so they are always watching for these techniques, but it’s still a game of cat and mouse and sometimes dependent on configuration