Skip to content
PodSubs
ShowsPeopleGuestsJobsNewsEvents
More
BlogPodCred
Menu
ShowsPeopleGuestsJobsNewsEventsBlogPodCredNetworksSign in
Join nowSign in
  • Shows
  • People
  • Jobs
  • News
  • More
    • Guests
    • Events
    • Blog
    • PodCred
    • Networks
    Join nowSign in

Tech Talks With Kinsoft

Tech Talks With Kinsoft artwork

Last Week in Tech - Ivanti's Unpatched Chain, a Record 1,400-Fix Patch Tuesday, and the Essential Eight Gets a Successor

Oct 18, 202622 minAudio
ShareXLinkedInFacebookWhatsAppEmail

Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! G'day and welcome to Last Week in Tech from Tech Talks with Kinsoft. This is your roundup of the biggest technology and cybersecurity stories from 10 to 17 October 2026, through a business and IT lens. Ivanti Connect Secure zero-day. Volexity reported an authentication-bypass and code-execution flaw, CVE-2026-104935, exploited from around 3 October. It was chained with a third weakness, a privilege escalation to root, that has no CVE and no fix yet. Ivanti says a fix is due in the coming weeks, as of the end of last week. A record Microsoft Patch Tuesday. Roughly 1,400 vulnerabilities fixed on 13 October, with experts pointing partly to AI-assisted discovery. We look at what that volume means for patching rhythm in a small team. Cisco Secure Email Gateway. A flaw reported as exploited, with fixes that shipped a month before the advisory and no fix at all for the 16.0 release line. ASD Essentials. The Australian Signals Directorate has launched the successor to the Essential Eight, moving away from the maturity-level ladder. Ahold Delhaize. The retail group, with more than 9,400 stores, took systems offline after a security incident confirmed on 12 October. Cause and data impact are not yet known. AI and the cost of fixing bugs. Bank of America's CEO described 1,000 contractors brought in to patch vulnerabilities surfaced by AI, and an Australian wealth firm has been penalised $1.35 million over cyber security failures. Visit www.kinsoft.com.au to talk through your security and IT needs. Sources: Volexity; Rapid7; The Hacker News; Dark Reading; CSO Online; BleepingComputer; Cyber Daily; ITnews; CNBC via Benzinga and ITPro; ASIC media release 26-244MR.

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 you just bought like the most expensive, heavy duty, absolutely impenetrable deadbolt on the market for your front door. Right. Top of the line. Exactly. You install it, you lock it, and you feel completely safe. But what if the lock itself, I mean, the very piece of hardware you bought to keep intruders out is actually the exact mechanism. welcoming them inside. Yeah, that completely flips the concept of physical security upside down. You know, you stop worrying about someone breaking a window and you realize

the primary threat is, well, the premium door you just installed. It's crazy. And welcome to today's Deep Dive. We are unpacking a massive tech and cybersecurity roundup today, covering the week of October 10th to the 17th, 2026. It's been a busy week. Oh, heavily. Our mission... for this deep dive is to decode a sudden, frankly, terrifying spike in cyber vulnerabilities. We need to figure out why our digital front doors are under siege right now. Yeah, and we're also

going to look at how the entire paradigm of corporate security is shifting. It's moving out of the IT backrooms and, frankly, directly into the boardroom. Exactly. It's no longer just a headache for the tech support desk, right? No, not at all. I mean, it is a foundational issue dictating whether a business can even... you know, open its doors tomorrow morning. Right. So whether you're running a company, managing a team or just trying to navigate a world where your personal data constantly feels up for grabs, we are going to connect the dots for you today. Because it really does feel precarious right now. It really

does. We'll get why it feels that way and what the new rules of survival actually are. And to understand the current threat landscape, we have to start where the attackers are starting. Right at the perimeter. Yeah. They aren't trying to break through the walls anymore. They are walking right up to the edge devices. And those edge devices, they are the literal digital front doors. We're talking about appliances like VPNs or email gateways that sit on the absolute perimeter of a corporate network. Right, the stuff facing the public. And security firm Valexity just disclosed

a highly critical zero -day flaw in Avanti Connect Secure. Yeah, CVE -2026 -104935. That's the one. It's an unauthenticated bypass that leads to remote code execution. So think of a VPN as the bouncer at a very exclusive club, right? I like that analogy. Yeah. This bouncer is supposed to check your ID, verify you're on the VIP list, and only then let you into the corporate network. Right. A good bouncer keeps the chaos out. But this vulnerability means someone on the open internet with no username, no password at all,

can just stroll right up to that bouncer. Just walk right up. Exactly. and essentially mind -controlled them into running their own malicious commands. Wow. Yeah. The unauthenticated part is what really keeps security professionals up at night. Attackers do not need stolen credentials for this. They don't have to guess a password or anything? Nope. They don't have to trick your accounting department with a clever phishing email. They just scan the internet, find the appliance, and take over. Which means if the bouncer's compromised, all those fancy security cameras inside the club are basically useless.

Completely useless. And the timeline on this Avanti flaw makes it even worse. I mean, Flexity found it being actively exploited in the wild around October 3rd. Yeah, early October. Right. But fixed software builds like version 25 .1 .1 .2 weren't published until roughly 10 days later. That is 10 stray days of a wide open door for anyone who knew how to turn the knob. And honestly, 10 days in the cybersecurity world is a lifetime. During that window, an attacker group tracked as UTA 0576 was... Incredibly busy.

Oh, I bet. Yeah. And they didn't just use this single flaw. They actually chained three different weaknesses together. Chaining. That's where they use one flaw to get to the next, right? Exactly. It's a fascinating, if terrifying, tactic. They use one minor flaw to get a tiny foothold, which then positions them to exploit a second flaw. Which gets them deeper. Right. And that gives them the exact access they need for a third flaw. It's like... Picking the front door lock, which gets you close enough to the alarm panel, which you then short circuit to open the main safe.

Perfect way to put it. And the reports note that while the first two weaknesses in this chain, the authentication bypass and the remote code execution, those are fixed in the new updates. Okay, good. But the third weakness is not. Wait, really? The third one is just left open. Yeah. That third one allows a privilege escalation straight up. to root access. And as of the end of this reporting week, there is absolutely no fix for it. That's insane. So patching doesn't even fully solve it. Well, patching the system closes that initial front door, you know, stopping

new attackers from casually strolling in. But if a sophisticated group is already inside, that third unpatched flaw means they can escalate their control to the absolute highest level. Root access. Root access means they own the machine. They can do anything they want. Oh, wow. This is exactly why Valexity explicitly warned businesses not to just rely on the vendor's built -in integrity checker. Because it might miss it. Exactly. The automated tool scanning the system can actually miss this specific complex chain of attacks completely.

So, I mean, if you're an IT director listening to this right now, patching just became the bare minimum. The absolute bare minimum. The assumption has to shift from we applied the update, we can relax to we might already be compromised. Like if your appliance was exposed to the Internet in early October, you basically have to treat it as breached. You do. And that requires a completely different operational mindset. It means manually digging through server logs for subtle signs of compromise. Which takes forever. It does.

And crucially, it means you have to rotate all the credentials that flow through that device. Oh, because they were watching. Exactly. If an attacker was sitting inside your VPN for a week, they were quietly watching all your traffic and harvesting all the usernames and passwords going past them. Which raises a really provocative question from our sources. Does the VPN even need to be facing the open Internet at all anymore? It's a great question. Right. Because if the bouncer is the biggest liability, maybe it's time to stop putting the bouncer out on the public street corner. And network architects are debating

that exact question right now because, you know, it's not an isolated incident with Avanti either. There's a similar highly critical flaw in Cisco's secure email gateway. Oh, right. CVE. 202676519. That's the one. And in that case, an attacker can write arbitrary files and run commands as root without even logging in. Again, unauthenticated. Yep. But what makes that Cisco flaw so strange is the timeline. Cisco shipped the fixed software releases way back on September 14th. Okay, September.

But the advisory actually warning customers about the flaw didn't go out until a month later on October 14th. Wait, why the delay? They only sent it out after they learned it was being actively exploited in the wild. So the vendor built a new lock, but didn't tell anyone the old lock was dangerously broken until people started reporting break -ins. Pretty much. It highlights a massive breakdown in communication between the vendors building our defenses and the companies relying on them. That is just wild. When the people making the locks are telling us after the fact that

the locks were broken the whole time, you really start to worry about the rest of the house. You absolutely do. We aren't just seeing targeted attacks on these edge devices either. We are drowning in a sheer overwhelming flood of software flaws across the board. Like Microsoft's October patch Tuesday just broke records. Oh, it was a massive drop. Yeah, they fixed roughly 1 ,400 vulnerabilities in a single day. Yeah, and around 80 % of those were rated critical. Just for some context, the previous month, September, saw about

974 fixes. Jumping to 1 ,400 is an enormous leap in volume. Huge leap. And that batch includes a mark of the web bypass, right? CVE -2026 -96951. It was actively being used by attackers. Yeah, that's a nasty one. For anyone listening who might be unfamiliar, the mark of the web is that little digital flag Windows attaches to a file when you download it from the internet. It's the mechanism that triggers that annoying but... helpful pop -up warning asking, are you sure

you want to open this? Right. The one everyone just clicks through anyway. But bypassing that flag means a malicious file downloaded from a shady website can masquerade as a completely safe local file. So it doesn't even warn you. Exactly. It strips away your chance to second guess the download, tricking the operating system into executing it without any warnings at all. Wow. But the bigger question here is why we are suddenly seeing 1 ,400 bugs in a single month. I mean, the sources point squarely to AI -assisted vulnerability discovery. Oh, AI is completely

changing the game here. Right. We have artificial intelligence models literally hunting for flaws at superhuman speeds. It feels like we are frantically bailing water out of a leaky boat. That's a good way to look at it. If AI is finding 1 ,400 holes in our hull every single month, are we just bailing faster and faster instead of building a better boat? I mean, that is arguably the defining tension in modern cybersecurity right now. We are supercharging our ability to find flaws, which on paper sounds fantastic, right? Sure. You want to find them

first. You always want to find the vulnerability before the bad guys do. But the logistical burden of testing and deploying 1 ,400 patches across an enterprise network, it's crushing IT departments. Teams literally need a collective vacation just to recover from Patch Tuesday. I can't even imagine. And the cost of that burden is just staggering. I was reading the Bank of America CEO's report, and they said they spent a couple of hundred million dollars plus last quarter on software fixes and anti -money laundering efforts. Hundreds

of millions. Yeah. They had to bring on 1 ,000 contractors, and that is on top of their 3 ,000 existing staff, just to patch the flaws being surfaced by AI models like Anthropix Mythos. A thousand temporary workers just to keep the water from sinking the ship. Right. It just shows that finding the bug is only the very first step. Testing the patch to make sure it doesn't break your existing software and then deploying it across thousands of machines, that takes massive human capital. But hold on, is AI actually helping us here or is it just generating expensive busy

work? Well, it depends on who you ask. Because the source material includes an interesting counterpoint from Mozilla's CTO. They mentioned an AI found 271 bugs in the Firefox browser, but they claimed there wasn't a single one that an elite human researcher couldn't have found. Right. Human experts are still top tier. So is this technology making us safer or just creating an impossible to -do list that costs hundreds of millions of dollars to process? I think the key... Is that AI is scaling the discovery process, not necessarily

inventing new categories of vulnerabilities. Okay, so just doing it faster. Yeah, it is doing the tedious, meticulous legwork of an elite human researcher, but it's doing it millions of times a second, 24 hours a day. Wow. So yes, it is definitely creating an impossible to -do list. And this is exactly where organizations have to shift their strategy toward prioritization over panic. You simply cannot patch everything at once. But how do you even begin to prioritize 1 ,400 red alerts? I mean, if my dashboard is

completely lit up, where do I even start? So the source cites Tyler Reguli from Fortra, and he brings up a crucial, honestly calming statistic. Of all the Microsoft vulnerabilities discovered in 2026, fewer than 1 % are actually known to be exploited in the wild. Wait, really? Less than 1 %? At the time of that report, it was only 26 out of the thousands found for the whole year. Out of thousands of terrifying red flags, only 26 are actually being used by attackers

to break into systems. That's it. That's it. And that is the secret to not burning out your team. If you try to patch all 1 ,400 equally, systems will break from untested updates, and you'll likely miss the truly dangerous, actively exploited flaws hidden in all that noise. That makes sense. Yes. You get lost in the weeds. Exactly. The lesson is to patch the actively exploited ones first and anything facing the open internet first. You have to build a sustainable rhythm of patching rather than treating every single AI -discovered bug as an existential crisis.

Okay, but when that rhythm fails or a patch doesn't hold, the consequences spill out of the digital realm and into the physical world almost immediately. They absolutely do. Look at the real -world chaos of the October 12th security issue at Hold Del Haze. For context for our listeners, this is a massive retail group. They own supermarket chains like Albert Heine, Food Lion, and Stop and Shop. They're huge. Yeah, we are talking about more than 9 ,400 physical stores globally. And operating at that scale requires an immensely

complex digital nervous system, tracking inventory, payroll, logistics, and customer data in real time. Right. And while the physical stores managed to keep their doors open during this incident, That digital nervous system took a massive hit. A very public hit. Yeah. Employee posts indicated that staff couldn't access their schedules or their pay stubs. European broadcasters reported that websites and apps went down completely across multiple brands. It was a mess. Now, our sources are very careful here, and we need to be too.

We have to be completely impartial. Ahold Delahaye's hasn't confirmed the exact cause. Right. They haven't said much. They haven't stated if it's ransomware. and they haven't confirmed if any customer data was stolen. We are strictly looking at the symptoms of a major outage, not diagnosing the disease. And treating the symptoms, as reported, is really the only responsible way to analyze an ongoing incident like this. But those symptoms teach us a vital lesson about modern digital dependency. How so? We have optimized our businesses

so much that we completely forgot how to function without the network. Oh, absolutely. I mean, picture walking into your workplace tomorrow morning. and every single core digital system is just gone. Terrifying. The screens are blank. No schedules, no payroll access, no internal communications. It's like relying entirely on your phone's GPS for a cross -country road trip, and suddenly you hit a dead zone in the middle of nowhere with no signal. Yeah. And you just freeze. Right. Do you still have a paper map

in the glove box? And even if you do, do you even know how to read it? Or is it some dusty relic from 2010 that no one understands anymore? Every single business leader needs to be asking themselves that exact question today. The core business lesson from the Ejol Delez situation is the absolute necessity of the offline fallback. The offline fallback. Yes. True security resilience isn't just about building higher digital walls to prevent hacks. It is about continuity of operations. The walls will eventually be breached. A patch

will eventually fail. The servers will go dark. So what does that offline fallback actually look like for a massive modern supermarket? It means having manual analog procedures ready to go at a moment's notice. If the scheduling software dies, can your store manager print a roster on actual paper? Right. If the point of sale network goes down, can you still process a customer's payment using an offline terminal or cash? Can you issue a physical payslip or a manual check so your employees can still pay their rent on Friday? Wow. Going back to paper checks. The

operational technology of running a business has become so deeply intertwined with the IT network that when the network falls down, the physical business often freezes completely. Resilience means knowing how to decouple them in an emergency. And the stakes for getting that wrong are incredibly high. We are talking hundreds of millions of dollars in mitigation costs and massive paralyzing operational chaos. It's enough to bankrupt a company. Definitely. And because the fallout is so severe. Governments and regulators are stepping in and completely changing the rules

of the game. The era of cybersecurity being written off as just an IT problem is completely dead. Oh, it's buried. We are witnessing a total paradigm shift in corporate governance. The liability is rapidly moving up the chain of command. Right. Because when businesses are paralyzed and customers are locked out, governments eventually drop the hammer. And we're seeing that regulatory hammer drop right now in Australia. Yes. Big news out of Australia. On October 15th, the Australian Signals Directorate launched ASD Essentials.

This framework replaces their older Essential 8 maturity ladder. I remember the Essential 8. Yeah. The old system was basically a standard checklist moving from level one to level three. The new system is described as a principles -based approach. Which is a big shift. Massive. It focuses heavily on identity management and, tying perfectly back to our first segment, the security of edge devices. But most importantly, it explicitly demands defined executive and board accountability. It completely removes the plausible deniability

from the boardroom. It connects technical failures directly to leadership oversight. But let me push back on this for a second. A shift from a clear level one to three checklist to a vague principles -based approach makes me a little suspicious. Doesn't moving to principles just create a blurry line? It can seem that way, sure. I mean, it sounds like it just makes it easier for regulators to point fingers and hand out million dollar fines to executives when things inevitably go wrong. Because a principle is so

much harder to define than a simple checklist. Well, you have to look at it from the perspective of why checklists fail in the real world. Checklists inevitably turn cybersecurity into a bare minimum compliance exercise. OK, I see what you mean. An IT director checks the eight boxes, hands the piece of paper to the board. The board glances at it, says, great, we're secure, and files it away. They did the bare minimum to pass the audit, and they never think about it again until the next audit. They treat it like a chore rather than a dynamic defense. Exactly. Principles force

genuine, continuous risk ownership. You cannot just check a box for identity security once a year. You have to prove. constantly that the principle of secure identity is being upheld across your ever -changing, evolving network environment. That makes sense. And if you want proof of this new regulatory teeth, just look at the recent ASIC action against Fortnum Private Wealth. Oh man, this is a wild story that should make every executive sit up. The Australian Securities and Investments Commission ordered this financial advisory network to pay a $1 .35 million penalty.

plus $250 ,000 in costs. That's a massive hit. Huge. And this was over cybersecurity failures that ended up affecting about 55 ,000 clients between 2021 and 2023. But the real kicker here is that ASIC openly admitted the company's conduct was not deliberate. Right. And that ruling is the ultimate proof of the paradigm shift we are talking about. You do not have to be malicious to be held legally and financially liable anymore. Just negligent. just have to be negligent in your digital defense. Regulators are now looking

at weak cyber controls not as a stroke of bad luck or the unfortunate result of a sophisticated, unstoppable hacker. They are treating a weak defense as a direct legal breach of your obligations to your clients. So throwing up your hands and saying, we got hacked, we were the victims too, is no longer a valid legal defense in the eyes of regulators. Not if you left the digital front door unlocked and the alarm system turned off. The expectation now is that the CEO and the board of directors must own the cyber risk just as seriously and with the exact same level of scrutiny

as they own financial risk, market risk or workplace safety. It's just the core business function now. Exactly. It's a core business function. Well, let's pull all these threads together for you. We started at the physical front door, looking at the deadbolts that are actually welcoming the intruders in. Those are the vulnerable edge devices like VPNs and email gateways. which are under relentless attack by adversaries who are chaining flaws together and hiding in the blind spots. Then we explored the sheer overwhelming volume of the threat. AI models are supercharging

the discovery of vulnerabilities, creating a massive financial and logistical burden on businesses trying to patch everything. And the key survival tactic there is prioritizing the active threats over the theoretical ones. Yeah. And we also saw what happens when the digital layer fails. with massive retailers facing physical gridlock. It highlights why every single organization needs an offline fallback plan. You have to know how to read the paper map when the GPS suddenly cuts out. Absolutely. And finally, because of all this escalating chaos, regulators are completely

changing the rules. They are tearing up the old IT checklists and holding the boardroom legally and financially accountable for cyber negligence. The overarching theme of this entire week of news is that passive security is no longer an option. Organizations have to actively manage their perimeter, understand their true vulnerabilities in real time, and take absolute leadership ownership of the risk. Which leaves me with one final forward -looking thought. We talked about massive banks using AI to find vulnerabilities and then deploying

a thousand humans to patch them. But if AI is accelerating the discovery of flaws to a volume that humans, even well -funded armies of humans, can barely keep up with, Will we soon reach a tipping point? It feels like we have to. Will we soon reach a point where software updates, security patching, and active defense are entirely managed by AI, removing human oversight completely? And if we get so frustrated with a heavy -duty deadbolt on our front door that we decide to

just hand the keys over to an artificial intelligence to lock and unlock it for us, what entirely new, unpredictable risks does that create for our future? Yeah, that's the million dollar question. It really is. Something to think about the next time you lock your front door. Thanks for joining us on this deep dive.

Open the transcript file

Hosts and Guests

Nobody is credited on this episode yet.

Credit the people on this episode

Be the first to rate this episode

More episodes

All episodes
  1. Iris Ohyama - A Suspicious File in the Cloud, 7.7 Million Customers at Risk, and an Online Store That Went Dark

    Nov 19, 202620 minAudio

    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...

    Credit the people on this episode

  2. Who Gives A Crap - A Marketing Supplier Breached, Six Days to Tell the Client, and Customer Data That Didn't Need to Be There

    Nov 17, 202619 minAudio

    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...

    Credit the people on this episode

  3. Last Week in Tech - An Outlook Zero-Day That Leaks Your Password Hash, Chrome's Eighth Exploited Bug, and Nvidia Without Jensen Huang

    Nov 15, 202621 minAudio

    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...

    Credit the people on this episode

  4. Chainalysis - From "Non-Sensitive" to "All Users": How a Breach Notice Changed in Five Days

    Nov 12, 202619 minAudio

    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...

    Credit the people on this episode

  5. NSW Health - A Dental Voucher Website, Four Gigabytes Taken, and the Providers Whose Details Were Exposed

    Nov 10, 202623 minAudio

    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;...

    Credit the people on this episode