this post was submitted on 22 Jan 2024
575 points (97.8% liked)
Technology
57472 readers
4632 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 1 year ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
Obligatory: companies should face harsh penalties for this stuff.
They do, in the EU. If you fuck up your customer's data, you'll face fines consisting of hefty percentages of your yearly revenue!
https://www.enforcementtracker.com/
Yep, hefty. Top 5: 1.2B meta, 746M amazon, 405M meta, 390M meta, 345M tiktok (all in €).
Oh noooo, 1% of their yearly gross revenue or 1.3% of their yearly gross profit. What a fine!
Side note: I would love to discover a public record of them paying these fines... we hear they ate fined, but never that they had to pay them. What is stopping them from cutting a deal of a payment plan over 20 years with 0% interest or full up front but only paying 30% of it or some lobbying BS.
We can infer that for sure this fine is coming out pre-tax.
This is not something a company did.
The group of people took a list of user names and passwords from a different breach and tried them on trello to see if people used the same password and wrote down which ones did.
Nothing a company can possibly do to stop this, only users can.
Even if the company required 2 factor authentication to fully log in, getting this far would still confirm each account/password combo was correct, which is all the "hackers" did.
That's not what happened.
Attackers queried n email addresses against trello, who responded with names and user names for accounts that existed.
No one asked trello to publish their names, so that's a breach.
This isn't completely true, but it is the current standard.
A website can detect and block many user/password attempts from the same IP and block IPs that are suspicious.
Websites can detect elivated login fails across many IPs are react accordingly (It may be reasonable to block all logins for a time if they detect an attack like this)
I'm sure there are other strategies, I don't know how often they are actually employed, but I wish companies would start taking this sort of attack more seriously (even if it's not at all hacking)
CGNAT would throw a wrench in that when you have thousands of users using mobile data and they appear to be coming from the same ip.
Nooooo, people keep telling me IPv6 will be insecure because of no longer having NAT.
Mostly people who don't know what a subnet is, but people.
You look for trends, not raw numbers. If an ip increase 500%in 10 minutes... throttle it a bit... insert wait times. If it's trust worthy then allow new value to become normal... otherwise keep the ip throttled.
That would be a P1 incident and probably violate SLAs depending on the duration.
Inserting a literally meaningless delay like 5 seconds is sufficient to make your service virtually impenetrable to mass bruteforce/stuffing attacks. Credential stuffing become untenable when your trying to stuff 1million creds with a 5 second cooldown. Most normal users who would hit it would just think their wifi or cell service hicupped.
I mean, passkeys are a thing.
Yes but this wasn't a data breach. This was a data stuffing incident, meaning they took someone else's data dump and tried their email and credentials here.
It's a breach.
Attackers queried email addresses and trello responded with names and user names.
real names is definitely a breach
Oooh that's pretty bad
Physical token over TOTP authenticator?
all the root secrets are available in plain text the generator app at some point, they have to be. moving that to a single purpose device greatly reduces the risk of vulnerabilities in your phone leading to exfiltration via internet connection
I cannot think of a use-case outside of statecraft. Maybe companies engaged, or being engaged, in corporate espionage.
Do you own a Yubikey?
Have you ever succeeded in getting it to work with anything??
It didn't work with gmail, or any other online account I had.
An absolute waste of $$.
mine works for my personal google account, work one is sso and doesn't have it enabled. otherwise gh, aws, auh0 support it, I'm forgetting some others I use. beyond that you can generate 2fa codes too
Setting up: https://www.yubico.com/setup/yubikey-5-series/
Supported services: https://www.yubico.com/works-with-yubikey/catalog/
Google Accounts (for your gmail): https://www.yubico.com/works-with-yubikey/catalog/google-accounts/
I use yubikey everywhere it's available for me. Initially, the first few websites in the early years were challenging. I think a lot of devs were still trying to figure out the workflow.
But today, it's usually as simple, or simpler, than TOTP.
So it might be worth trying again. I'd use a YubiKey 4 or higher if you can. If you have an older one, you may want to upgrade to take advantage of the newer technology like NFC and Bluetooth if you're into that.
I just wish YubiKey could store more than like 30 TOTP tokens.
Sounds like a skill issue.
Have had yubikey for a few years. It was a pain to set it up initially, but it took me less than an hour if I remember correctly. Since then the only issue I have is that sometimes I accidentally bump into it and it pastes an OTK to a random place.
I use mine with AWS.
tbf it's just email, username and real name so it's basically nothing when half of users are [email protected] either way.
For project tools like Trello, a good portion of your userbase is company emails. A malicious actor now has a list of company emails that they can compare against public facing data like Linkedin, imitate a user using a gmail based off their name, sending an email to that company's IT team asking for an MFA reset sent to the newly created gmail account. Now imagine if that compromised user is a developer with admin access to production environments. These were the conditions for various ransomware attacks.
An email, username, real name are not much, but it's a foot in the door.
It is a foot in the door but honestly there are way too many doors out there so it's really hard to measure the real damage of this.
I worked at a pretty major employment company like 20 years ago when basically everything was legal and we didn't need to buy dark web datasets to find real names and contacts ever - most of that data is publicly available and can be captured with simple public scrapers and email checks.
I think expectation of names and emails being private should be thrown out of the window entirely and every security system should implicitly assume these details are publicly known.
So the conditions I mentioned were directly from a series of ransomware attacks from the group BlackCat including the high profile ransomware incident targeting MGM Casinos last year. My team recently used the same premise during an incident response drill based on that event.
I agree that data security is important, even if it is only email addresses, where many are probably findable in the web anyway. Maybe, the link with the username has some value, but I’d bet only little. In my opinion, harsh penalties are more needed in privacy invasive (in my opinion malware) like google, meta, Amazon etc. are spreading.
The problem is that this data can be combined with other data. An email address by itself isn't particularly important but when it's matched up with names, physical addresses, DoB, SSN, other PII and the network of other services with matching data it becomes very serious.
It's never just this breach, it's every other breach as well. Every breach makes every preceeding breach more effective and more valuable.
Except this contains none of that
Other breaches do.
If two breaches have an overlap, e.g. they both contain email address, then they can be joined into a more complete set.
Yeah, I don't think there is much that would be gleamed by combining with this dataset
Of course, but where are names, physical addresses, DoB, SSN, etc in this dataset? It’s just mail and username
Other breaches do.
If two breaches have an overlap, e.g. they both contain email address, then they can be joined into a more complete set.