Hacker News Reader

Comment by gonehome | Hacker News Reader

This twitter thread is pretty bad and the comments here aren't much better.

IIRC the Ubiquiti 'hack' was an insider attack from an employee lying and intentionally breaking things while pushing his lies to the press to hurt his employer. Krebs was wrong and tricked by the employee. I don't know if that justifies this legal action, but it's not the normal going after someone who reported a breach. This one is more complicated.

I'm pretty sure Corey is wrong on the facts in this case (and so was Brian). I also felt a lot better about Ubiquiti once the dust settled and the details about Sharp came out.

Edit: I missed this comment thread which basically says the same thing: https://news.ycombinator.com/item?id=30850793

View on HN

You have a loosely defined term: "bad." It suggests "inaccurate" but perhaps you mean "unkind."

In either case, you assert "Krebs was wrong" a couple times but you skirt the topic of the twitter thread, the comments, and the lawsuit: that Krebs intentionally misled readers in order to drive traffic for ad revenue.

Do you think Krebs did this?

Krebs got the big story exactly right: Ubiquiti had a bad security problem and flubbed the handling of it.

Krebs certainly did not get the story “exactly right” nor did he take responsibility for how wrong he was (and how he was used by the attacker).

Maybe? You aren't presenting any evidence for "how wrong he was" and "how he was used by the attacker." You're sure of it? Why? What other explanations are possible here? None?

There was a lot of discussion at the time this stuff went down: https://www.theverge.com/2021/12/1/22812761/ubiquiti-data-br...

The HN comment majority also changed from when I made my first comment. At the time I was pushing back against the overwhelming sentiment here and on Twitter that Ubiquiti was going after a journalist just for reporting a breach. The story is way more mixed than that and from the stuff I remember reading at the time Brian wasn't in the right.

I don't want to be overconfident (I don't really know the specifics), but I did want to push back against the overwhelming overconfidence I saw arguing the other direction.

Yeah that's fine but just stating things like that as incontrovertible fact with absolutely no supporting evidence, none, which is what you did, isn't useful. It made me pretty suspicious of an astroturf for example. 9 times in 10 it isn't that but it always looks really bad. It's also tantamount to saying "everyone here is either an idiot or acting in bad faith" and has no reason at all to believe a thing that you are now pointing is is wrong and so clearly wrong that it's the same as stating water is wet. I note now you're providing evidence you're walking it back a bit, which shows the worth of doing so.

It's very common elsewhere. Just keep repeating a thing until it is accepted but it really isn't how things are supposed to be in sensible discussion.

I think that's an unfair characterization.

I was replying to someone who said Krebs was "exactly right" - that's what I was pushing back on. My first comment was also more nuanced. The evidence exists for those who care to look and I think what I wrote is inline with it.

If you'd provided any support with that statement at all, or said /why/ you believe that, it's completely fine. Doesn't even have to be correct! We all get stuff wrong. If you believe you provided the support in a different comment, cite it.

"I'm right, go find the evidence if you dare doubt me." Not so useful.

If that standard is applied equally that’s fine with me.

Well we're trying, huh? You and me both and I appreciate it.

> Krebs got the big story exactly right: Ubiquiti had a bad security problem and flubbed the handling of it.

This can be true while still being massively exposed to legal liability. The court of public opinion is one thing, the law is another.

(Except in the US, the law is very biased toward allowing people to say stuff on matters of public interest, and the legal liability from the complaint appears to be limited to paying a lawyer to make it to, and file for summary judgement. https://anti-slapp.org/virginia)

Regardless of his reporting at the time, Krebs has known since December when Sharp was arrested that:

- the supposed hacker was the goddamn Unifi Head of Cloud, using the access keys needed to do their job

- the initial internal investigation into the hack and ransom was being conducted by the attacker

- that the whistleblower account is a complete fabrication by the internal attacker and his reporting on a coverup are false

Ubiquiti aren't suing him for reporting on it, they're suing him for not retracting it properly once it was revealed how false it was. As per the filing:

70. Ubiquiti brings this litigation because of Krebs’s refusal to do the right thing and retract the March 30 article or the December 2, 2021 update, which continue to malign Ubiquiti’s reputation, damage its relationships with its stockholders, and disrupt its business operations.

Krebs and Corey are _way_ wrong on this.

There is plenty to be said and very valid criticisms about how Ubiquiti dropped the ball and handled the situation. The attack being an insider does not excuse them. But it invalidates much of the reporting.

Krebs was specifically and personally targetted by the attacker as a method of spreading false statements to damage the company and, by keeping the articles up, remains complicit and liable.

It might be more convincing if all the things you listed don't make ubiquiti look really bad for a company that needs you to trust them, given their gear is very cloud-y.

From my perspective this lawsuit looks like they've Streisand Effected the fact that they let their internal security be even worse than the initial accusations.

It's like finding out your financial advisor had all your money stolen, which is bad enough, and then it turns out it's because they gave their gardener the password to your bank account.

Where's the proof that Krebs knew that the whistleblower account was a fabrication?

Genuine question.

It seems to me that Krebs might be in a position to claim:

1. He honestly reported the facts as were made available to him.

2. Either:

A. He didn't know his original source was Sharp (quite feasible that Sharp disguised this)

B. He did know his original source was Sharp, but felt compelled to continue to protect his source despite charges having been brought (innocent until proven guilty).

3. He took the view was that nothing in the revelation of this hack as an inside job casts doubt on his initial reporting, which was about Ubiquity's response to the incident not the attacker's identity.

We should at least wait to read the defence before drawing any conclusions.

He honestly reported the facts as were made available to him.

Which is a more than adequate defense. See New York Times vs. Sullivan.

The truth is an absolute defense against libel, but what happens when that stops being true?

The parent comment referred to Sullivan v. NY Times, which ruled that a public figure (and I suppose companies count as such) have to prove not just factual error, but actual malice.

Seems like refusing to take down known inaccurate articles at is at least suggestive of actual malice.

That appears to be Ubiquity's argument, but we've only got one side at the moment, and to my reading (uneducated as to US defamation law) it's far from proven.

Well yes you wouldn't expect anything to be proven at this point. There is only a complaint. We are discussing the reasonableness/plausibility of the allegation. All I'm saying is that to me it seems plausible.

Again, "actual malice" doesn't mean "acting maliciously" in the colloquial sense. It's a legal term of art from the aforementioned NYT v. Sullivan case and is explicitly defined as "with knowledge that it was false or with reckless disregard of whether it was false or not." Publications leave up articles they know to contain falsehoods (that were believed to be true at the time they were written) all the time. Unless libel (and therefore actual malice) is shown, I don't know of any precedent that would imply an obligation to issue a retraction.

Erm, I took a quick read of the December 2, 2021 update ...

https://krebsonsecurity.com/2021/12/ubiquiti-developer-charg...

What part of that update is incorrect? Naming that update is not going to help their case, at all.

This lawsuit is likely doing the exact opposite of what Ubiquiti expected.

Before the lawsuit, I had some sympathy that they got jerked around by an ex-employee with major access and took it in the shins. I'm kind of in the glass houses and stones camp ... I doubt very many companies could withstand a high-level technical person going rogue. They found the problem. Now they're pursuing charges against him and that's rattling through the legal system. Sure, there's lots of reputational damage, but that's the kind of thing that happens when you centralize management of things--it makes them a high profile target (see: Solarwinds).

However, the lawsuit against a reporter is causing me to pause and think "Wow. Maybe they're actually institutionally incapable of recovering from this, worried that something else might get exposed and really do suck."

The lawsuit moved me from slightly sympathetic to Ubiquiti into "What kind of idiots think this is a good idea?" and looking for alternatives.

There's valid criticism to be said about the corporate structure and culture which allowed this all to happen, we're all in agreement on that, but the lawsuit (while probably not accomplishing exactly what they'd hoped for) is legit if you ask me, in the sense that a journalist must take responsibility for the stories they put out and the informants they trust.

I'm not a journalist so I don't know how these things are supposed to go, but shouldn't Krebs have verified his sources identity before publishing? Isn't that a thing journalists are supposed to do?

That said I think Krebs was right to publish the story at the time, but when it became clear that Adam actually was Sharp the story should have been retracted. Perhaps Krebs should even issue an apology at that time?

> the supposed hacker was the goddamn Unifi Head of Cloud, using the access keys needed to do their job

If this comment from a former employee is correct then no, he had root access to a bunch of stuff for no good reason and their security stance is abysmal.

Nobody should have the root aws tokens. They should be split between two teams and stored in a safe & access should go through another method that is audited

https://news.ycombinator.com/item?id=29456593

The employee in question was the head of their cloud, so he would have been the one to implement, or drive the implementation of the proper access controls. Based on other employees accounts of the guy, it sounds like people were trying to advocate for better access controls/separation but he didn't let it happen (presumably because he was planning on doing something like this).

I've been on the other side of something similar to this. At a previous role, a security researcher was falsely claiming we had a backdoor. We considered litigation, but ultimately decided not to for a variety of reasons, but a major one was that there was no way that the optics would be anything other than "softare giant sues researcher," and we would likely only serve to draw more attention to the claims.

I'd entirely forgotten about the Ubiquiti breach until today.

> Krebs has known since December when Sharp was arrested that:

Criminal Accusations <> Facts until proven in a court of law. All Krebs knows at this point is Sharp was arrested.

And what we all don’t know at this point is whether Ubiquiti is competent enough to have unfettered access to all their customer networks because they failed to defend against insider threats.

Is it even possible for a company to defend itself from sabotage by the person who is presumably responsible for their security? Seems illogical.

They can design devices that are resilient to attacks by even themselves. That's hard when your whole pitch is "your devices are connected to the cloud!" though. But maybe that's the problem.

> Is it even possible for a company to defend itself from sabotage by the person who is presumably responsible for their security? Seems illogical.

Off the top of my head:

a) don't force your clients to add their networks to be accessible by your cloud, this was their entirely huge mistake. or by design to enable spying activities. the same way I can log into unifi and set a switch port in promiscuous mode and forward the traffic to my remote ip, so can they.

b) two people required for secure access to sensitive systems.

c) sensitive gear on-premise under constant video surveillance.

d) logging to remote servers under control of "internal security" not "security", regularly monitored by "internal security".

Companies do this and more. They do it by contemplating a solution to a problem rather than dismissing the solution as "illogical".

Remember "Apple’s insider-trading policy enforcer accused of insider trading"?

He got inside information from being...chairman of the committee that managed disclosure of sensitive information.

https://arstechnica.com/tech-policy/2019/02/lawyer-who-wrote...

But shouldn't good journalists verify the identity of their sources before publishing stories?

> IIRC the Ubiquiti 'hack' was an insider attack from an employee lying and intentionally breaking things while pushing his lies to the press to hurt his employer.

"Their undetected security breach was by an insider" is not the sterling defence of Ubiquiti that you seem to think it is.

These types of attacks are extremely hard to defend against.

Considering the nature of the attacker (and the HN comments about the guy at the time of the attack) my take away of ubiquiti from this event and their response was positive. I also down ranked my expectation of Krebs’ accuracy.

It’s not personal - he writes lots of great stuff, but his response to being wrong in this case was worse than Ubiquiti’s response to the incident.

Its really not that hard to defend / audit the actions of a single lone wolf, even if they're the head of security.

This is just more and more embarrassment for a company that clearly just doesn't understand security, and yet wants to force all their users of their products to rely on them as a trust model by making all of their ongoing product systems completely cloud reliant or at least holding third party root keys to sell ads / customer telemetry.

But the only reason the insider job worked is because they insist on making their gear dependent on the "cloud" for no good reason?

Their response, which was to not disclose the breach? Or their late response, which was a vague email you might need to change your credentials? Or the fact that their cloud environment was a joke run by a one man show?

It's like Okta. Of course it's hard to protect against an insider ( although Ubiquti didn't really try), but that's not an excuse to screw up the disclosure.

> These types of attacks are extremely hard to defend against.

And yet, enterprise-level suppliers do it successfully every day.

It was more nuanced than that. There were extortion attempts, abuse of trusted access, deliberate attempts to mislead and misreport.

How many developers that you know have been fired for introducing security bugs?

The insider was responsible for their AWS architecture. This isn't a lowly disgruntled employee.

Security incidents are still security incidents regardless of who was behind it. Putting it in scare quotes doesn't make it any less real. Why should paying customers feel better about any of this mess?

It doesn't help that Kreb's whole shtick is security alarmism. Of course he's going to take the bait, he's over eager.

Reply on news.ycombinator.com