KDDI – One Flaw, Six ISPs, 14 Million Exposed Mailboxes
18 minAudio
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! A single vulnerability in one shared platform cascaded across six Japanese internet providers, exposing up to 14.2 million email accounts. Telecoms giant KDDI has confirmed a breach of a shared email system used by KDDI, JCOM, NIFTY, BIGLOBE and others - an intrusion via a flaw in third-party software. We look at "shared-infrastructure blast radius": how concentration risk turns one weakness into millions of victims, why exposed email logins are gold for credential-stuffing, and the transparency questions KDDI has left unanswered. Do you know how many of your services depend on one shared platform? Visit www.kinsoft.com.au to talk through your security and IT needs. Sources: BleepingComputer; SecurityAffairs; The Japan Times.
Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.
Transcript
From the show. Tap a time to play from there.
So imagine this. You wake up on a random morning in June 2020 slugs. You check your phone and you find out your email provider has been completely compromised. Which is pretty much the worst way to start your day. Right. It's awful. Your stomach just drops and you know you immediately start doing the mental math of like every single thing connected to that email address your bank your shopping accounts everything. Oh yeah. The panic sets in instantly. Exactly. But. Here's the real shocker of this specific incident. For a huge
portion of these victims, the security vulnerability that exposed their entire digital lives didn't even come from the provider they actually signed up with. Yeah, that's the wild part. It came from a completely different brand, like one they might never have even heard of, just sharing the same invisible basement. It's a profoundly unsettling scenario, honestly. And it really forces us to confront this reality of where our data actually lives. We just assume that when we buy a service from company A, our data stays safe within the walled garden of company A. Right,
like it's locked in a vault with their logo on it. Exactly. But we don't realize that company A and company B and company C, they're often just different marketing storefronts. They're built on top of the exact same vulnerable foundation. Which brings us to what we're talking about today. Welcome to today's Deep Dive, where we are really going to unpack the architecture of that shared foundation. We're looking at a massive cyber incident involving the Japanese telecom giant KDDI. An absolute behemoth of a company. Yeah, tens of billions in revenue. And we're drawing
from this really fascinating incident analysis by Tech Talks with Kinsoft. So our mission today is to kind of decode how a flaw in an unnamed third -party software system managed to expose up to 14 .2 million accounts. And we really need to look closely at the mechanics behind that exposure because... I mean, a number like 14 million, that doesn't just happen by accident. No, definitely not. It is the result of compounding systemic vulnerabilities. And just to set the stakes here for everyone listening, this isn't just some story about a random tech glitch, right?
It's a massive lesson in what we call concentration risk, the hidden dangers of credential reuse, and honestly, why the exact wording of a company's public apology might be the most crucial part of a data breach. Oh, absolutely. The PR spin is always telling. Right. OK, let's untack this. Let's just lay out the timeline and the scale first. So late June 2026, KDDI confirms this massive breach. that they actually detected on June 17th. Yeah, right in the middle of the month, they spot this unauthorized access. Right, and
the attackers guard in by exploiting a vulnerability in some third -party software that ran a shared email system. And the ultimate fallout here is that email addresses and passwords for up to 14 .2 million users were exposed. Which is just a staggering volume of data. It really is, and that includes, you know, active users, former users, and accounts that had just been sitting dormant for years to kind of wrap our heads. around this it's like well think of living in a massive ultra secure high -rise building okay I like this analogy right you think you're safe
but then you discover the contractor used the exact same master key for all six high -rises on the entire block that is a brilliant way to visualize it because that's exactly how enterprise software architecture works today But, you know, what's fascinating here is that from a pure incident response perspective, KDDI actually did a few things right. Wait, really? Yeah, I know it sounds crazy given the numbers. Because exposing 14 million accounts sounds like a colossal failure on literally every front. It does sound like
it. Yeah. But the Kinsoft analysis points out that when they detected the compromise on the 17th, they actually acted decisively. They successfully blocked the attacker. They stopped the ongoing exfiltration of data. Oh, wow. OK, so they didn't just let it bleed out. Exactly. And they didn't try to sweep it under the rug internally either. They immediately began notifying Japan's privacy regulators and they reached out to the other Internet service providers who were affected. I mean, their containment protocols actually worked. Right. But containment doesn't undo the
fact that 14 million records are already gone. Precisely. The core problem that makes this so terrifying isn't their response time. The alarm bells are ringing because of the sheer scale of the initial exposure, the blast radius. Right, the blast radius, which leads perfectly into the first major structural issue from the Kinsloff report, this idea of concentration risk. Because that massive number, it isn't just because KDI has a lot of customers. Not at all. It is entirely due to that shared master key architecture you mentioned earlier. In cybersecurity, we talk
constantly about blast radius, like how far does the damage spread when one single component fails? And in this case, it spread really far. Unbelievably far. Because that shared email platform wasn't just for KDI's direct customers. It was the silent engine sitting behind six completely different ISP brands. Six different brands. So it's KDDI plus... Let me check the list here. JCOM, Nifty, Big Lobe, Chubu Telecommunications, and STNet. Yeah. And to the average consumer, those are competitors. Yeah, totally separate companies.
Exactly. So imagine you had a terrible customer service experience with KDDI five years ago. You get mad, you cancel, and you intentionally switch to NiftyY, thinking you're moving your data to a safer, separate infrastructure. Right. You're like, take that, KDDI. You're right. But underneath the distinct branding and the different websites, you are literally just moved to a different server rack in the exact same backend environment. Man, that is frustrating. So what does this all mean for you listening to this deep dive right
now? Think about your own workplace. Think about the digital tools you rely on every day. We all, you know, we love the efficiency of shared infrastructure. We rely on it completely. Yeah. It's why we use centralized cloud providers and universal single sign on. But how many of your own business operations quietly depend on a single shared component? Like what happens the day that one piece of third party software just fails? Well, as we see here, the blast radius multiplies exponentially. And this is entirely driven by the business economics of tech consolidation. I mean, from a CFO's perspective,
it makes perfect financial sense to centralize your back -end systems across all your subsidiary brands. Oh, sure. You cut costs. Exactly. You reduce your server overhead. You only pay one IT maintenance team instead of six. You streamline your updates. I mean, you are building this incredibly efficient machine. But efficiency always comes with a hidden tax. Always. And the hidden tax of efficiency is extreme fragility. By centralizing everything to save a few bucks, you have effectively subsidized the attacker's return on investment.
Subsidized the attacker? Think about it. Instead of a hacker having to figure out six different ways to breach six different companies, they only have to find one single flaw in that shared third -party software. And once they break down that one door? They get the data for all six brands simultaneously. We are building these massive, highly optimized digital towers, but we're balancing them on single points of failure. Which is terrifying. But exposing that shared server isn't even the end of the story. Because the attackers, you know, they aren't just sitting around looking at a spreadsheet of 14 million
email addresses. They immediately weaponize that data. Which brings us to the fuel for the fire here. credential stuffing. Yeah, because an email address and a password combo, it's not just an isolated piece of information. It is literal currency in the cyber criminal underground. I think a lot of people just assume like, oh, my telecom provider got hacked, whatever. The worst they can do is read my spam folder. Right. Or look at your old phone bills. Exactly. But the real threat has absolutely nothing to do with the telecom provider at all. Not even a little
bit. Attackers do not care about your old emails. What they care about is human psychology. They know that we all suffer from extreme password fatigue. Oh, I certainly do. We all do. You're asked to create an account for everything. Your bank, your pharmacy, your streaming services, the place you order pizza from. And as a result, the vast majority of people just reuse the same two or three passwords across their entire digital life. Here's where it gets really interesting. It's like using the exact same key for your front
door, your car, your safe, and your office. So a thief... steals your key ring from, say, a hotel valet. The thief doesn't care about your car in the hotel parking lot. No, they don't care about the hotel at all. Right. They take that key and they walk down every street in your hometown just hitting the unlock button over and over again until your front door at home suddenly clicks open. The vulnerability isn't just the email leak. It's the human habit of password reuse. That is the exact mechanism of credential stuffing. But you have to realize
it is happening at a scale and speed that is almost impossible to comprehend. When attackers get a data set from a breach like KDDI, they don't type them in by hand. They feed those millions of username and password pairs into sophisticated automated bot networks. OK, wait, I have to jump in here because don't most modern websites have defenses against that? Like if I type my bank password wrong three times, I get locked out or I have to click on, you know, all the pictures of crosswalks in a CAPTCHA. How are the bots getting past that? That is a great question. And it's where the economics of cybercrime get
incredibly advanced. The hackers know all about rate limiting, that defense that locks you out after three tries. To bypass it, they use something called rotating residential proxies. Basically, they route their bot traffic through millions of hijacked home internet connections all around the world. So to your bank's security system, it doesn't look like one hacker in a basement trying 10 ,000 passwords. Oh, wow. It looks like 10 ,000 normal people. Exactly. It looks like 10 ,000 regular customers sitting in their living
rooms trying to log in exactly one time. They fire these stolen credentials at corporate VPNs, healthcare databases, Amazon accounts, all simultaneously. So they just completely circumvent the basic security tripwires. Entirely. And if we connect this to the bigger picture. This exact mechanism is why enterprise IT leaders are so aggressively pushing for zero -trust architecture. It is exactly why you, the listener, are constantly dealing with security friction at work. You mean the endless nagging to use an authenticator app?
Yes. The nagging has a very specific critical purpose. This credential stuffing pipeline is why enforced multi -factor authentication MFA is simply no longer optional. Right. Because even if they have the password. Even if the attacker has your exact password from the KDDI breach, if they do not have that temporary six digit token generated on your physical phone, that stolen password is mathematically useless to them. That makes so much sense. It's also why relying on human memory for passwords is just
a fundamentally failed security model. Strict password managers and never reusing a password. Those aren't just for tech geeks anymore. They're the only way to structurally break the chain. OK, so we see. how the ghost kitchen architecture allowed the breach to happen in the first place right and we see this devastating economic machinery that takes the stolen data and weaponizes it but um there is another layer to the story that i found incredibly revealing when reading the kinsoft analysis and it happens right after the dust settles ah yes when the company actually has to face the public Yes, the crisis communication.
Because the way a company words its breach disclosure might actually be the most vital metric we have to judge their internal culture by. Let's look at KDDI's actual public disclosure. When they announced the breach, they included the specific line noting that, quote, some of the compromised passwords were hashed or encrypted. Right. And they crucially omitted two massive pieces of information there. First, they completely refused to state which specific cryptographic algorithm they used. And the second, they refused to clarify
what percentage of the passwords fell under the category of some versus how many were just left completely unprotected in plain text. Now, wait, let me play devil's advocate for a second here, because when I read the passwords were encrypted, my immediate reaction is relief. It sounds reassuring. If a password is run through a hashing algorithm, it's scrambled into random gibberish. So even if the hackers stole the whole database, isn't the data essentially useless to them anyway? Well, this raises an important question, and it completely exposes the danger of vague reassurance. Because not all encryption is created equal.
In fact, saying your data is hashed without naming the algorithm, it's the cybersecurity equivalent of saying your front door is locked without mentioning that the lock is made of wet cardboard. Wait, really? But a hash is a one -way mathematical function, isn't it? Like you can't just unhash something. In theory, yes. If KDDI used a modern computationally expensive hashing algorithm, something like bcrypt or Argon2, along with a unique randomized salt for every single user, then you're totally correct. It would take forever
to crack. Right. It would take an attacker an absurd amount of computing power and literally decades of time to guess those passwords. We don't actually know if they use those modern algorithms. Because they wouldn't say. Because they wouldn't say. And a lot of older legacy IT environments, which, by the way, are the exact kind of environments that get rolled up these massive corporate consolidations companies, are often still using obsolete hashing algorithms like MD5 or SHA1. And those are bad. They were designed decades ago to be fast, which is the exact opposite of what you want for password
security. You want password hashing to be slow and expensive. Okay, but even if it's an older algorithm, if it's still a one -way function, how do the hackers actually crack it? Oh, they don't even bother cracking them manually. They use something called a rainbow table. A rainbow table. Yeah, because older algorithms like MD5 compute so incredibly fast, hackers have already pre -calculated the hashes for literally... billions of common passwords. Right. Like they already had these massive lookup tables sitting on their hard drives. Yeah. So if they steal a database
of MDFI hashes, they don't compute anything. They just compare the gibberish in your database against their pre -calculated rainbow table. It takes a matter of seconds to reverse engineer millions of passwords. Wow. Okay. So if KDDI used an outdated algorithm, the fact that they said the passwords were hashed offers basically zero actual protection against a modern attacker? Correct. And by withholding the name of the algorithm, KDDI is actively preventing security professionals and their own users from accurately calculating
their true risk level. It is the difference between genuine transparency and just, you know, PR damage control. Which brings us to that second omission, which honestly feels even more deceptive to me. Using the word some to describe how many passwords were protected. Oh, in statistics and cybersecurity, some is a terrifying word. Right. Because if some means, say, 99 percent of the 14 million accounts were strongly encrypted and maybe 1 percent slipped through in plain text because of some weird logging error. I mean, that's a
problem, but it's a contained problem. Exactly. But some could also mean 10 percent. Yeah. If some means only the newest accounts created in the last year were hashed. Yeah. And 90 percent of the legacy database was just sitting there in plain readable text for anyone to download. I mean, that is a catastrophic failure of basic security hygiene. So by using this vague language, they get to sound super reassuring to the press while essentially leaving their own users completely in the dark about how fast they actually need to be changing their banking passwords. Because silence matters immensely in cybersecurity. Transparency
isn't just about issuing an apology. An apology does not secure a network. True transparency is about providing actionable technical intelligence to the victims so they can protect themselves. Vague reassurance actively erodes trust because it signals to the industry that you are either hiding severe negligence or worse, that your own internal IT teams don't actually know the extent of the damage yet. Specific honest detail, even when the news is objectively terrible. builds trust. It empowers the customer. Okay, so as
we pull all of these threads together, let's recap the real journey of this deep dive. Because the KDDI breach of June 2026, it is so much more than just a story about a software bug. It really is a masterclass in evaluating the blast radius of shared infrastructure. Exactly. We've seen how that relentless drive for digital efficiency breeds massive concentration risk. We've explored the cascading dangers of credential stuffing and how our bad password habits basically fuel the fire. And finally, we've seen why the absolute necessity of transparent, specific crisis communication
is so critical. It's a stark warning, really, for everyone. it is and there is incredibly actionable advice here for you listening right now the kinsoft analysis strongly suggests taking time to map out your own single points of failure look at your vendors look at your software stack where is your concentration risk quietly hiding. It's always hiding somewhere. Always. And if you need a starting point for that kind of audit, Kinsoft actually offers great resources for mapping this stuff out at kinsoft .com .au. Their core mantra
for this is really one we should all adopt. Stay patched, stay skeptical. And I would just add a quick point regarding consumer empowerment. When a service provider you use is breached, and let's be real, statistically speaking, you're going to be caught in one of these eventually, do not settle for the boilerplate PR email assuring you that they, quote, Take your security seriously. I hate that phrase. Right. You are fully entitled to ask planted questions. Demand the technical truth. Was the data encrypted? What specific cryptographic algorithm was utilized? Were the
passwords salted? Don't let them hide behind vague words like some. Demand the specifics so you can accurately gauge your own risk. Which leaves us with a final lingering thought to ponder as we wrap up today. We've seen that our digital economy is entirely driven by consolidation. Sharing servers. Centralizing identity access. Rolling multiple brands into single back -end platforms. forms, right? It all drives operational efficiency. But as we continue to centralize more and more of our digital lives into fewer and fewer massive interconnected hubs, are these
cascading catastrophic failures simply the unavoidable tax we must pay for modern convenience? Or as the blast radius of these breaches continues to grow, is it time to start intentionally building a little bit of inefficiency and isolation back into our systems?
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! Japanese home goods company Iris Ohyama (online store operator Iris Plaza) found a suspicious file in an external...
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! Australian online retailer Who Gives A Crap told customers on 7 November 2026 that their information was accessed...
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! Last week in tech, 7 to 14 November 2026. Microsoft's November Patch Tuesday (10 November) fixes well over a...
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! Chainalysis disclosed on 30 October 2026 that an external actor using "offensive agentic techniques" accessed...
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! NSW Health reported unauthorised access to its Oral Health Fee For Service Scheme website on 29 October 2026;...