you glossed over this point, but it is nonobvious. how exactly does making your users skeptical of email "lose the war"? are they going to type in their passwords out of spite now?
the gp is exactly right. if you have a line you wont cross to train your users, then youre leaving a vulnerability. if you want to make the argument that the vuln is worth not "being mean" to your employees, then make that argument, but don't pretend its more secure.
No, they’re going to quit, or put less effort into their job (which is often worse for the employer than quitting), or tell other people not to work there, or—as in this case—tell everyone on the internet what you did and then your company’s reputation takes a well deserved hit.
This is not a good scam. People fucking notice when they expect a 650 dollar bonus and it doesn't come. It's not some arbitrary "check out this cool link" style email that an employee might click and then forget they've ever engaged with.
Not to mention that it caught nearly 10% of the fucking company - That's an insanely high click through for any phishing test I've ever seen at a software company. And I don't even know if they sent it to all 7k employees (my guess is no). So either GoDaddy has incredibly incompetent employees, or this looked fairly real (and the email does actually look pretty damn good for a phishing email - custom image, no spelling mistakes, internal domain, aware of different geos the company operates in)
So you pissed of 10% of your company with an email that wouldn't ever be sent as a real scam (because a week later the employee is going to ask where the 650 bucks is and alarm bells will go off).
I don't have the context of a Godaddy employee, so maybe there's an obvious mistake in there, but general speaking - this was a dumb play by the security team. It's now also losing them goodwill in public.
Phishers can easily get what internal company emails look like. Send an email that looks like "Hey, we're giving you a $900 bonus. Make sure to fill out these forms on workday", with a link that is a Phishing website that looks exactly like workday.
Fill in your credentials - hey, also need your 2factor code - and bam, hackers have an in to the system. It doesn't matter that the employee realizes 10 days later they were phished.
Security is like an onion - It's layered into place across different levels, and different systems.
If you just want access to a single employee's machine, for a short duration - sure this would work. If you really want to compromise the whole internal net, it's going to take longer than a few hours to work your way into other systems. Generally you either need your target to access the system you want, or you have to spread to other machines to find a machine with the access you need.
The most effective attacks are the ones that get in, and have weeks to spread through the whole internal network. Take the Target breach in 2013 - They were in the system undetected for 20 days, nearly 3 weeks.
As soon as the company knows they have a problem, any decent security team is going to check every system on record.
In the 10 days or 1 day that it takes between realizing they were phished, all sensitive information they can get access to can be stolen. Furthermore, more sophisticated phishing links can then be sent from their account. After all, who's going to suspect an actual email send by a colleague as a phishing attempt?
A holiday bonus type of phishing attack absolutely can work, and be extremely effective at credential theft. It may not be effective at literally scamming money from the employee, but who cares.
Know what doesn't build better security culture? Trying to trick your users. Know what does? Working with them closely to help them understand security, finding out when and how people get tricked, and working to solve those issues.
The anti-phishing efforts I've seen so far have been lame and ineffective. Rather than trying to find new ways to make people fail, security teams should be finding new ways to prevent people from falling victim.
I don't think GP was ambiguous at all, to be honest. This point was extremely clear. It feels like you came to the discussion with an axe to grind.
And people will treat you like that, most often behind your back.
How about a phishing test that downloads something that nukes your hard drive?
> How about a phishing test that downloads something that nukes your hard drive?
The phishing weakness opens that possibility. The act of wiping the HD does nothing to improve behavior or identify weakness.
They didn't feel they had any choice but to hire about 50 temp folks from a local contracting company and run through the entire desktop environment and pull drives, re-image, reinstall. They did this over the course of a weekend and on Monday when employees came back they found their desktops wiped clean. Within an hour the site lead started getting calls and visits from panic stricken employees that had kept all sorts of personal info on those machines...photos, emails, documents, etc.
All gone.
They ended up missing one machine that got powered up that Monday and re-infected the entire campus. Had to do the same thing again.
The fact that employees are failing this proves the point, and until we can show that they are no longer falling for it, it shows how vulnerable we remain from an infosec posture.
The only reason you would have to send something like this is because you are lazy, need to generalize quickly, and blast it to the company. And you will teach a lesson. And you will kill any goodwill to the security team as a byproduct.
That said, I do agree with you that money is better spent elsewhere. A company with > 1000 employees should just assume there's an infected machine in their network 24/7. Defense in depth stipulates that this should not matter for their security.
I also prefer my security measures procedural, technical and layered. Like I said before, phishing simulations don't really work for me at all... Currently many places where I do some work the security department have a large say in what happens, regardless of goodwill. Security measures always suck, are less efficient and cause no end of headaches, you mostly have 0 goodwill anyway.
Isn’t your job to make sure security measures don’t suck, aren’t less efficient, and don’t cause headaches?
This seems to be a very defeatist attitude to take.
From my personal experience one of the following happens when a security team takes your perspective: 1.) users work around security mitigations, causing worse security issues 2.) workers quit the company due to friction when working 3.) the security manager gets fired because they won’t ever compromise
You seem oblivious to the fact that you can still be effective but not a shitty security team.
Your way of approaching this literally makes people who should listen to you and trust you want to do the opposite. What an awful security team you’re part of.
It's funny you say that my way of working literally makes people want to do the opposite of trust me, when I send them a phishing e-mail that's exactly what I'm aiming for ;P
I've also worked at a place where the security team were trusted collaborators. Devs were comfortable communicating with them. Their security skills improved over time, and so did the security of the software they wrote.
The latter strategy is far more effective at moving the needle over the long term.
Treat. People. Decently.
I said shitty. No other swearing and no name calling. What exactly are you reading?
The real thing happened, was detected and shut down. Then they did some trainings over the course of a year, and had us test the same thing again next year. Results were not particularly good, and people got upset that they did not receive the promised Christmas gifts.
Afaict, these trainings stop being effective at around 10%. That is, 10% of all people being phished like this will do everything you tell them to. Up to and including sending your their password and installing malware on their machines following your e-mail instructions.
I agree this shouldn't need to be a "gotcha stunt" though but if this would be to you then yes you're doing it/communicating it/following it up wrong. It should be an awareness call that phishing is going to target you in very thoughtfully real and meaningful ways, not just boring fake tickets/issues or the like. If you can't find a way to send that message in your org then again I claim your org already has bad culture as it can't talk about or examine real security issues and instead has the first reaction to a real risk that you're just trying to improve your reporting numbers again.
Also a small year end bonus being not only considered exciting but not existing at all is probably the real callousness problem in the org, not that one points out the allure of money is used by malicious actors to get people to click things. But that's separate from the whole role of security discussion.
At the end of the day if you've got a threat vector you're too afraid to actively measure it doesn't matter how much you message other departments about it you'll never know how big of a risk it is or if appropriate action was taken unless you are compromised by it a later date. This is true of human risk monitoring as much as it is of technological risk monitoring.
1) An uptick in phishing emails (both user reported and data analyzed) was noted with a few patterns of attack in particular being on the rise in the last month.
2) Communications about the attack profiles were sent out company wide followed by invitations to material (written and short video form optional for low risk groups in the org based on use of email and HR training courses for those with heavier/job required use patterns that fit the attack profile, usually IT).
3) About a month time pass, user based reporting numbers were looking better and communication to leadership that a test email was going to go out to users matching the attack pattern to see what click through rates would look like on a well tailored email that wasn't quickly removed from inboxes. Leadership approves.
4) Click troughs result in dropping at an internal landing page hosting links to the above communications and adds the HR training course to the users profile.
For more detail on step 4) the opening of the landing page is a forgiving style not a scolding style, think "Oh no, this could have been from one of the phishing attacks we've been experiencing lately. Don't worry, it was just an example attack from us - but did you notice any of signs of a...".
It's also worth noting as part of 2) it was discussed the official communications on end of the year packages would be sent out _before_ the security test, not sure if that was the case in GoDaddy's scenario. This year didn't include cash bonus at Christmas (we did one in the summer for COVID) but did include unlimited PTO rollover and similar non-cash perks given the situation (and explained there would be no cash bonus).
3) Can be tricky when trying to fend attacks on the leadership level, it usually uses a modified approach to this list where lower management of each department is involved instead.
As far as the numbers themselves the higher the click through rate the WORSE the score the security group gets, it's a score for the amount of improvement from the planned action not a score for how many people you could trip up before you get a pat on the back. At the same time it was modeled after the COVID related phishing emails we had been getting, not something someone made up because it seemed easy or hard to pass.
It's a very large org and in the end this went well for all teams (even though improvements weren't quite as good as we had hoped they were still pretty darn good numbers) and we were able to show with data we had improved security against a real threat. I'm sure there were a few (given numbers) that were a bit let down after clicking it but given the approach and planning I don't think that was a failure of the test approach (though we're always open to refinement when an opportunity is brought up. Part of the training material is freeform feedback on how you think the org could better handle the challenges from your PoV).
.
Anyways the point is it shouldn't have to be "tackle real risks or have other groups hate security", you need to find a way to do both at your place (which may be different than how to do both at another place). And that should be true regardless if GoDaddy managed to implement the test poorly this time or not.
The most good-faith reading of this whole tire fire that I can manage is that if you're going to pull this, you had absolutely and without question better have actual bonuses, of no less and frankly probably more than the phishing attempt, in the pipeline for every employee.
If you don't, you should be fired because it is inhumane to act this way to other people. It is, and I do not use this word lightly, evil.
What is bad about a company culture where going for maximum emotional impact over a weighted approach causes an uproar? (I'd be much more wary about one where it doesn't, because it tells me employees expect the company to yank their chain, and expect bad treatment from other teams, and know complaining about it won't help. that's a broken environment.)
I disagree that one can characterize the simulation of a specific scenario as "lazy." Is it unusually attractive to the target? Yes. Lazy? No.
Specifically, you would have to prove that sending phishing training emails with more neutral topics (e.g. a package arrived, IT policy change - ACTION REQUIRED) is less effective than sending the more potentially harmful.
Something like: “SECURITY INFORMATION: Phishing emails target holiday bonuses to increase engagement. Always be on alert” along with a few points on what you are likely to see in a phishing email, how you could spot one, and what to do if you get phished.
Are you sure about that? Is there any evidence that supports this claim?
It certainly fails the sniff test. Behavioral science does not hold setting you up for failure as an effective learning contingency. I mean try teaching your dog that way, see how far you’ll go. You will also teach your workers that phishing emails come from internal so there is no need to report it.
When I was working as a life guard we never went into training situations unknowingly. There is a reason for that, a) it is dangerous as workers might act irrationally creating a dangerous situation, b) it is stress invoking and not healthy for anyone, especially those that have underlying stress issues or conditions, and c) there is no evidence that live training is any more effective or teaches you anything more than traditional training does.
Result at my work. Email is completely ignored by most everyone. All communication is via Slack, even when not appropriate.
I check my email about twice a week, and by the end I’m usually browsing job boards out of disgust.
But these emails presumably weren't even coming from the company's domain, since they were phishing emails.
Go look at the image in the article.
I don't work there so maybe something in it should have set off red flags for employees, but generally speaking - That's a very high quality phishing attempt.
> Sent by [email protected]
Invoking and then revoking a bonus email in a year where folks are already hard hit financially and mentally under the guise of “education” is lazy and any security leader who does it is a fool.
Effective security culture builds rapport with business units. This type of nonsense does the exact opposite.
Yes you may have proven a tactical point, but you have just set yourself up to lose the war.