This includes who talked to whom and what the responses from the manufacturer, supermarkets, and cold store warehouse operators were.
Don't expect to learn what the vulnerabilities themselves are, though. They couldn't get hold of and trace a subset of the device users, so they're holding back what the vulnerabilities are.
Interesting the company that may have manufactured some or all of the units, claimed that physical access to the devices would be required.
Perhaps, you can trigger a hard reset on the device and the firmware returns to default network password?
> Resource Data Management (RDM), the British manufacturer of the refrigeration controllers, disputed the severity of the identified vulnerabilities. The company argued that exploiting them would have required physical access to the hardware.
I read that claim as "They need to get one physical example of the hardware" and that this would them allow them to attack all instances of that same model.
Which, to me, reads like a secret stored on the hardware. Something like a hard-coded password, storing a single private key on all devices (or a public key with a crazy weak private key), or some other symmetric secret used somewhere in authentication.
Out of curiosity, how do you know that? I would imagine (and really hope) you are correct, and if you have insider knowledge of this investigation, that is great and I appreciate your added context here.
Not the target's hardware, no. The firmware is in update files and the units sell cheaply secondhand, so physical access to the device you're attacking isn't needed.
Physical access only comes into getting the firmware, not exploiting the device. That’s how these were found: pull the firmware off a unit you physically have, then analyse it. But the firmware isn’t a secret that needs the hardware. It’s in upgrade files and other places, so you can get it any number of ways, examine it, and find exactly the same issues. And the vulnerabilities are exploitable over the network. “You’d need physical access” describes one way we got the firmware, not a barrier to exploitation.
"RDM, though, says comparisons with the units attacked in mainland Europe are not fair because hacking their fridges and freezers would have involved getting hold of a piece of hardware.
The ethical hackers say this would have been achievable as the hardware in question can be purchased cheaply online on sites like eBay."
... TTL serial? CANbus? Some dumb service mode dongle?
I'm surprised the fridges need to have far-remote control over their temperature in the first place. I can absolutely see the need for an internet-connected thermometer inside them to be able to alert when things go wrong off-hours, but that's a read-only resource.
I can also understand the purpose of centralised temperature computer control through the local network, but that's not exposed to the internet.
An internet-writeable path seems like inviting disaster.
I'm not sure about this particular case, but I believe the intention is to support remote maintenance and repair. For example, if a fridge ices up, you need to raise its temperature and defrost it. Of course, in many cases, this involves emptying the fridge, removing the evaporator fans, and spraying hot water on the coils until all of the ice is melted, which is obviously not something that you can do over the internet.
You might also want control over the temperature to react to electrical prices or grid stability, but that's pretty niche and could likely be 90% scheduled (cool slightly too cold at night, then allow to get slightly warmer (while still food-safe) during day).
I agree though - this should be behind a management VPN, not just on the open internet.
I'm quite curious to know what the actual attack was. The article is light on detail but does quote 'similar to weak password' but admits that it would have been necessary to get hold of an example device first, which isn't that hard as they're very common.
To me this spells out an unprotected private key or secret residing in the device that could be used to break the authentication checks. In which case once secret is out there - no need to posses it? Could the key/secret have been well protected, but the device could be used to sign anything given to it?
On BBC Radio 1 during one of their news segments, they interviewed someone involved, who claimed it was as if the passwords didn't need the correct password, any password would be accepted.
Writing software as if it's the 80s? I wish! Aside from hobbyist code and games written for home micros, software of that period was usually extremely efficient (out of necessity) and comprehensively documented as standard.
-1 it's AI coming for our laziness - 1000s of CVEs the past month from well regarded software of all kinds, even Linux issuing waves of CVEs and PostgreSQL 19 delayed for the first time in years.
> It was thought the UK had largely escaped such hacks as the British refrigeration industry relies on different devices but there was in fact a potential vulnerability the sector has been rushing to fix.
> In late October, security researchers discovered a major flaw that, if found by hackers, could have led to even worse damage to the UK.
Having worked with industrial equipment like this that often had some kind of network interface retrofitted to a 20+ year old device; I am going to guess it was a manufacture default master password that can be used to remotely manage/reset/reprogram the device and that essentially all deployed units never changed/disabled it.
In fact it may not even have been possible to disable or change it. It is scary the number of appliances, both industrial and consumer, that have "secret god mode" reset passwords and also just happen to work remotely..
Now to defend the manufactures a little bit here these master passwords were originally used because these devices didn't have any kind of network connectivity at all when they were originally design back in the 90s (or perhaps even earlier).
They were there for an engineer to come along, plug in an RS232 or similar and they could use the master password to just fix whatever was wrong. Then over time either networking/internet functionality was added via an additional interface (such as a master control unit all machines were connected to and then that MCU had network/internet access but essentially acted like an unmanaged network switch and just passed on whatever it was sent) or directly to the system in a new model revision.
However nobody deals with the whole master password part and so you end up with a warehouse or supermarket freezer that is now on the internet and if you're lucky "protected" by CG-NAT (usually that is the best case, rarely do you see any real network security in these setups) with essentially a remote login that all use the same password.
Of course this is just a guess from my own experience but if I were a betting man that is where I would put my money.
As for the "security researchers discovered a major flaw" I would also guess this flaw has been known by the manufactures since forever but it was security through obscurity as few people knew of a master password or how to use it.
However most likely aided by LLMs some researchers just dropped in firmware blobs from industrial systems they started looking at after the hacked systems in France in the summer and the LLM did what they do extremely well and find these things.
This set off the alarm bells and the NCSC got involved as a national security concern and thus begins a quiet but rapid call to mitigate the issue before someone else discovers it.
The article does say (admittedly from the manufacturer) that to exploit it required getting a device and taking it apart, so maybe they did just dump this password from the firmware.
Perhaps I'm being a bit too charitable, but surely if the password is indeed just one single root master password, once it has leaked, its leaked for good and so there's no need to get hold of a device. Or someone could have brute-forced it to begin with...
This has got me wondering if my alarm system; which is connected to my network and has a cellular backup is a lovely hackable bridge with default password 0000.
Stupid question: is there any reason these supermarket fridges should go up to above zero degree? If no maybe the user can set a hard limit that can only be manually overridden in store? At least it gives them more time before the food spoils?
In my time, freezers built a growing layer of frost over time. When you felt your fridge had too little space left up there, you unplugged the device for minutes / hours and manually removed the thick ice plating before it melt.
That's why they put automatic defrost in fridges now. Connecting fridges to the internet, on the other hand....
I wish there was a "brick its wifi" AI-box. You setup the box, you select the target, it worms itself in, finds the wifi driver and bricks it. Or if it can find or make alternativ firmware. Just device reconqista without desoldering stuff.
what would this accomplish except DDoSing the target?
and these kinds of attacks already happen, consistently. e.g. the Polish Power Grid attacks where they wiped PLCs and disabled network access and observability systems.
and if you're suggesting just killing their wifi to save them then I don't think you understand what or why they installed these systems in the first place.
In the UK, I think part of that is because you're explicitly warned not to tell anyone when the specific agency reaches out to you. But there is definitely significant pro-active work done.
UK is also vocal opponent of russian war in Ukraine, and not just empty words but also actual supplies. Many important objects and folks were obliterated with say Storm shadow missiles.
Which means this stuff is matter of national security in active warfare situation and we know well russia doesn't mind performing acts in other nations that it itself considers acts of war. Which is also fair, russia effectively quietly declared semi-open war with entire western civilization quite some time ago and secret services / governments are all well aware of that. They see us as a mortal danger to their system, which is also fair since it is when talking about their mafia oligarchy.
My father was a “tovarăș”, his father was a “tovarăș”, that is because they were both Romanian Communist Party members and proud of that. Unfortunately I’m not a “tovarăș” because neoliberalism got the better of what had been left of communism by the late 1980s, and instead I’m left on a desolate forum like this one ideologically fighting Westerners whose wealth has been built on colonizing the world. Bleak as fuck.
The British government (and a fair number of the people) don't seem to understand why Russia would have a problem with NATO while all these NATO-aligned militaries are cheerfully sponsoring attacks on Moscow. So in fairness, they probably won't understand the link you're alluding to either. I can't work out if it is willful blindness or they're high on their own propaganda.
In that time, russia has poisoned british citizens on british soil
Invaded Georgia
invaded Maldova
stoked a war between Armenia and Azerbaijan
Invaded ukraine, twice.
Now, this would be fine from the UK's point of view because those countries are not in the UK sphere of influence anymore. After all the days of thin red lines in crimea are very much behind us.
The issue for the UK is that all these military actions are getting closer and closer to our trading partners. Russia has been very clean on their imperialistic expansion.
Sure spending money supporting ukraine attack moscow is unreasonable, if we just ignore that russia has spend ~100-600k of thier own men in trying to annex the entire of ukraine. And what is next after annexing ukraine? well its poland, hungary, slovakia & romania. All part of the EU trading block, and very much our problem.
Its also not like russsia is being innovative or new here. Its not the first time britian has been sending supplies and support, along with the french to stop the russians doing expansion.
When Putin's invasion army stalled 30km from Kyiv in March 2022 he could have just retreated and called it a day.
Had he done so I suspect the war would not have continued; the "disputed territories" would still be in an uneasy balance and the world would have moved on despite shooting down a passenger airliner.
Ukraine did not chose war, Putin chose to start and continue the war and is the reason we are where we are today. NATO & the West has been very restrained and not got directly involved in the fighting, Putin though has put N.Korean troops on the front line which is surely a much more serious escalation.
The cheering didn't start out of the blue. Not.connected to NATO either: a fair few non-NATO countries that had the bad luck of neighbouring Russia, and so being candidates for imperial 'pacification', are also cheering it. Rightfully so.
NATO is a defensive organisation, not an offensive one [1].
Ask yourself why the ex-soviet republics are all seeking to join (or have joined) an organsistion that exists for defence amongst democratic countries.
[1] Something Mango-Mussolini was reminded of when he thought NATO would suport him invading Iran.
I would like to know how this played out. I'd like to imagine a bunch of spooks turned up at Tesco HQ and asked to speak to their IT Security with directives from "top men".
I guess the reality is someone watching a dashboard scanning the controllers each day and sending emails till they disappear.
That's usually the CISO, and both the Cyber Governance Code of Practice and ISO 27001 assume that they'll be the person responsible.
Neither is mandatory in the UK, of course. We really ought to bring in some sort of statutory duty to report cybersecurity breaches and near misses, but at the moment we only require that for a relatively narrow set of essential services.
Since this is all hush-hush (Official Secrets Act and such) and we know next to nothing I'm going to assume that they're using the same shitty Danfoss software everyone else is using and they just coordinated the updating and forcing all supermarkets to take their fridges offline. Nothing more.
There's probably some Orwellian law in Blighty that gives them the power to command private commercial enterprises like that. That wouldn't be possible in the U.S. and other European countries. In most democratic nations the government can (in peacetime) only strenuously urge companies to update their software, not bully them around like this.
If you want to show that it never went off and the, e.g. medical or food product contents are therefore safe....and if there are a lot of them....
On the other hand perhaps they don't need to each have a connection individually to the internet but could perhaps be fronted by some device that's easier to update and keep safe.
As much as we're seeing stories of cyberattacks (if you can count using a default easily guessed and never changed password on a device a "cyberattack"), widespread system outages and a never-ending stream of CVEs, I believe this is temporary. We're unlocked a new capability and it's going to take awhile to catch up:
1. The OSs, major programs, languages and frameworks are all going to end up more hardened as a result of the bugs that are found and possibility exploited;
2. (Perhaps more importantly) The systems that companies are governments absolutely rely upon are going to have to deal with security and reliability in a way that they largely never have before. Such things were and still are treated as an unfunded mandate for which nobody is rewarded for fixing these things and nobody is punished by ignoring them. That era, I believe, is over. That's not such a bad thing.
> if you can count using a default easily guessed and never changed password on a device a "cyberattack"
Victim blaming. It's very easy to overpower an old lady in public, but it's wrong to do and it's a criminal act
I'm not saying that security shouldn't be improved, it should. But the fact there's a spree of opportunistic criminals because crime became easier for them to achieve doesn't make it not a crime
The problem is this is on the Internet. Yeah, a woman should feel safe walking around the city at night no matter how she's dressed up, and we can police the city so it's safer.
There's no doing that on the Internet though. The Internet has cryptocurrency and VPNs which means bad actors are really hard to catch.
No, victim blaming would be blaming store employees for not updating their fridge's security. The greater corporation is responsible for providing secure, reliable fridges that do what they're supposed to and making sure they're used correctly. By being negligent they're not victims, they're perpetrators.
It's no different to, say, (correctly) blaming a car company if their car under certain conditions can cause a steering lock or spontaneously catch fire when it's too hot outside. Blaming the victim would be blaming the individual drivers.
If a car maker puts poor locks on the door and ignition, and you break the locks and steal the car, that is a crime
If a car maker puts the strongest locks on the door and ignition, and the owner gets out leaving the car running and the door open, and you slip in and steal the car, that is also a crime
Stealing cars is a crime no matter how easy it is to do. Don't give me this "well it was easy for me so it's the car makers who are the real criminals" as you drive away having stolen my car.
The people breaking into fridges, regardless of the strength of the security, are criminals, and what they are doing is a cyberattack. It doesn't matter how easy it was for them or how negligent the fridge vendor was.
The corporation bought in from reputable vendor ... in your analogy they bought a car (fridge thermostats) from a car company (fridge thermostat company) that blew up on them
If they use a 'complex software/hardware' stack (open source or not), they are done for. We all know this is much more than fixing one flaw. Just a matter of attracting the attention of a nasty hacker group. Since it is from the UK, could be some news to just do that: honey pot (it is legal over there).
Is this… an extremely routine and ordinary security practice being turned into a news article? I read it and all I could get is “yes, they’re putting on their pants before leaving the house”. Will the bbc next run an article about git? “Losing work is a thing of the past - UK device programmers now record entire software history!”
Quite a few owners and operators didn't know these controllers were directly exposed to the internet. Smaller operators often had them on cellular modems and didn't know they were online at all. Some larger operators assumed they weren't exposed. They were wrong.
Then you have to work out who owns each one. There's the retailer and the refrigeration contractor. In many cases these were franchise stores. Seeing a device on the internet doesn't tell you whose it is. Geolocating the IP means very little when it's behind a cellular modem.
Then you have to contact them. "Your refrigeration controller is exposed to the internet and might get hacked, can you disconnect it?" is an odd phone call to get.
Then someone has to physically find it. Supermarkets don't have IT staff on site, and you don't send IT into a refrigeration plant room. That's travel, logistics and organisation, site by site.
And you have to know that disconnecting it is safe. If that controller runs the refrigeration for the whole store, how do you know pulling the Ethernet cable won't cause a serious problem? It all needed risk assessing.
This was a significant amount of work across a lot of organisations. It wasn't putting your trousers on before leaving the house.
And the damage inflicted in some European countries was limited. There wasn't any widespread food shortage or anything.
It was simply a result of negligence on the part of the manufacturer and supermarkets. Their claim that they make so little markup on food stuffs that they can't afford decent IT security sounds silly to me.
Security Researchers announce that BigCo leaves a key to the front door under the door mat!
BigCo denies that they do that, announces that security is their middle name and that they have corrected the problem.
Security Researchers announce that BigCo now leaves the key under the flowerpot.
BigCo announces that security is of utmost importance to them and have corrected the problem.
Security Researchers announce that BigCo now puts the key in an envelope labelled "Opening This Envelope is a Federal Crime! And there is no key inside!"
Security Researchers roll eyes.
BigCo sues Security Researchers, accuses them of terroristic acts and says that G*d wants them to make ever increasing quarterly profits and that they can't make money if they don't move fast and break things, and that they definitely did not use an LLM to create their security architecture.
Lmao try reading the comments on there, typical BBC HYS stuff. Buncha "hurr durr the world has gone to shit because of the internet and freezers on the internet!111" made by people who readily leverage said technology without realising it because they're all shortsighted apathetic NIMBYs.
But they're not wrong though, are they? Fridge automation shouldn't be directly accessible from the internet.
Directly sending monitoring/metrics? Maybe. Directly controllable? No! Absolutely not! At minimum, require a VPN tunnel to the fridges' network. They same should be true for anything that doesn't have to provide network services to the public.
Same goes for all IoT hardware (the S in IoT stands for security)
Is a vendor selling you a "smart" device, where the "smart" part happens only if you hook it up to the internet, and can be unilaterally changed at any time by the vendor because all the logic's on their end, not yours? Tell them to fuck off and put the "smart" on the device itself or you're not buying it, and while you're telling them that, tell them to fuck off with their subscription-based or subscription-alike business model, it's blatantly being done in the vendor's interest and not yours.
They may be morons, but the HYS crowd have entirely the right idea.
Can someone explain to me why on earth you'd want to put a fridge online?
For the first 150 years thermodynamics and refrigeration didn't need the internet, so why now?
Frankly I've no sympathy for those who have to have to manage this problem now. As I see it, this is yet another instance of using technology for it's own sake—just because it's now doable.
A national network of fridges uses quite a bit of energy to run. Cooling empty fridge space on that scale is a huge waste of energy and money.
Connecting them up to an integrated logistics system opens up big savings. You can start cooling only as a delivery is on the way and at temperature when it arrives.
Britain has historically had some of the best run supermarkets in the world. This is a way to pass savings on to customers.
It's one reason why fresh produce is generally cheaper in the UK than the USA even though the cost of energy is much higher in the UK.
No I don't think so. It's not so much just the reporting as the central control and coordination.
A local thermometer won't know that lory with an exact number of pallets is held up in traffic or what its ETA is.
An integrated logistics system can know where every lory is, what's on it and where it needs to go.
It can control the fridges for drop off along the way and can deal with changes in delivery schedule or on demand re-routing.
Imagine customer demand changes in one particular shop. In another one fridge isn't reaching temperature, re-routing a delivery will wastage whilst a servicing team is sent to fix it.
That sounds like a decent argument for being able to monitor all the temperatures, but that is a read only operation. In the hack stories I've read they were turning the fridges with food in them off remotely. What use is that capability? Turning off a fridge remotely when it is empty to save power? Couldn't that be done with a local sensor completely removed from remote control?
A local sensor could definitely turn an empty fridge off but it can't turn it back on again when you know the lorry is say 35 minutes down the road.
Without that the fridge will lag in getting down to the correct temperature for arrival.
I guess you could have drivers pull over and call into the shop with their location but I'm sure you can see that's open to potentially costly mistakes.
Because they have a warehouse of a thousand fridges that they need to be made instantly aware of if one fails and need to have an audit log of the fact that 'yes the food in there was definitely stored at temperature x at time y'. Multiplied across a country.
I get what you're saying but we're talking about industrial scale logistics and cold chain management. 150 years ago I wouldn't have been able to get whatever fruit I desired year round from the supermarket and honestly I like that I'm able to now and consider this real tangible progress.
Have you seen the cleaning rostra on the walls in public toilets, there's often gaps in that because people are people.
Also with the manual method it needs to be accurately written down, accurately entered, and then analysed to see if there are any problems building up, not just obvious failures. Having estate wide real time data can make a big difference.
There can also be a lot of data collected to predict problems. If you detect one unit using more energy to hold temperature then perhaps a component is failing and you can get it serviced before it breaks. They can also monitor efficiency, perhaps half of the fridges get the morning sun and struggle to maintain temperature so putting window blinds may help.
Question. What do you think of remote monitoring of computer servers, surely someone can just walk around the data centre, and your customer offices, with a pen & paper checking for warning lights. For example HPE SmartArray predictive failures of hard drives, who needs that!!
"Those systems were completely equivalent to the current internet systems,…"
Sorry, they were not. How could they be when TCP/IP still had to be invented? Moreover, many used HW that was physically incompatible with the low levels of the OSI model (the physical layers couldn't even connect).
Sure, but wardialling was a thing and these sort of systems absolutely were being hacked long before they were connected to the Internet. Even systems that were theoretically private (eg. multiple point-to-point frame relay connections ) would often have a modem connected at head office to allow on-call remote access out of hours.
A well designed modern network architecture is perfectly capable of providing a much higher level of security than any remotely-accessible system from the 80s or 90s.
The problem is the lack of that well-designed network architecture, not that modern technology somehow makes it inherently impossible to secure.
As a kid, I had a saturday job in a supermarket...... in the pub after work, we'd all be drinking away with our boss who would typically be "doing the freezer checks" and writing random temperatures in the log book at the pub table.
Indeed! I do appreciate it when someone brings up a story of how the world actually works.
"Who the hell wants to go peer at fridges all day and has enough incentive to not just sack it off" is my immediate response to those suggesting we can take a step to full employment by not having these things do it automatically.
The other thing not mentioned here is managing energy costs.
In the old days (say 20 years ago) energy was much cheaper at night, so to save costs, one of the things you could do is drop the setpoint of the freezers 10 degrees to extra cool the freezers, then set it back to normal during the day. This would save not insignificant amounts of cash. So long as the stuff didnt unfreeze, it doesn't affect the quality of the food (within reason, there are caveats, that someone more knowledgable can chip in with.)
Nowadays the the cost of energy is less predictable, but avoiding significant draw between 7-10am and 16:30 and 20:00 can half your energy costs. (https://agileprices.co.uk/?region=C this is customer variable pricing, but peak time is 2x the cost of offpeak, sometimes more.)
You don't need the internet for warning/failure detection, we had sophisticated systems in everything from power stations, water supplies/dams, burglar alarms etc. long before the internet. The intrinsic nature of these specialist networks made them hard to hack.
It's been known almost since the inception of the internet that putting systems online that used the standard internet made them highly vulnerable, so it's why critical systems should not be put online but use proprietary methods/protocols that cannot be easily hacked, especially so if used with specialized hardware built for the purpose.
It's clear to me this was either cheapskate design, or fashion, or done for convienience—i.e. check the status with a laptop or one's phone.
We've seen this security shortsightness previously, for example axing the job of the dam keeper who'd always be present to open and close dam sluice gates just because someone could monitor the dam online from afar—just to save a few bucks.
It's not as if these vulnerabilities weren't foreseen, they were. Now these dumb (and greedy) planners are paying the penalty.
Stores rent fridges as a service. Maintenance is done online by third parties. This has been going on for a decade or so, now the fridge still works, but no one maintains the software.
Putting fridges behind a VPN means an investment and installation in each physical store.
Now, some of these super markets are big chains, others are small owners with a franchise model.
That doesn't explain it. What kind of maintenance can be done to a fridge remotely? Even temperature monitoring doesn't seem justified. Just put dumb thermometers in each fridge, employ some dude to check them regularly and have a loud alarm if one goes out of range.
Right. Pre internet, we had good secure industrial remote monitoring systems that were essentially hack proof. Fashion and a desire to conform forced these systems onto the internet. It wasn't necessary (see my second post).
3,000 stores - 24/7 fridge checker - £100mil a year in cost per large supermarket to check to fridge temperature?
what happened to the hacker spirit that would see that as the basis for a business selling a subscription for chopped fruit in a bag you squeeze at home?
> Can someone explain to me why on earth you'd want to put a fridge online?
For example: So the service company can monitor temperature and get fault alerts and error codes remotely, shops can change temperature (for different foods) and internal lighting (when shop opening times change) without having to visit each unit on their hands and knees.
Cost savings by lowering the power consumption of hundreds of fridges in hundreds of supermarkets equals millions in savings.
I'm sure, buried in the details, is also uk.gov's weird and wonderful health and safety laws because of food standards, that somehow requires remote access to temperature control (e.g. 'all refrigerated food needs to be at -10.25C' - one click and you're done).
Honestly, that's overkill for this sort of application.
It'd be enough to have a bog standard segregated network with a basic security elevation mechanism, together with some reasonable amount of logging and monitoring. The sort of configuration that tech companies use to protect their database server.
Data diodes make sense in situations where you need security clearance to even be in the same room as someone who's accessing the high side - they're expensive and painful to use, and nothing a supermarket does is ever likely to justify it.
This article reads like it's from a parody cyberpunk universe - and still no-one asks the actual question: why on God's glorious Earth does a fridge need to be on the internet?
Want to know the temperature, remotely, for whatever reason? Put a 4-dollar sensor in it, and have it report to a read-only homeassistant-style setup. preferably behind a proper firewall.
Because, in the UK at least, the 16-18 year old Saturday kid, isn't trusted to change the temperature of fridges if asked.
And going with the 'one peanut less saves millions', at the corporate level, a bean counter probably figured out being able to remotely change the temperature of every fridge consistently across hundreds of supermarkets by a tenth of degree saves millions in power.
Being able to remotely read and set temperatures, diagnose faults, tell how much food is left, and generally do things at distance is incredibly useful for these industrial fridges.
Walmart internal propaganda stickers on the back of house refrigeration. "For every minute the door's open, it costs your store 55 cents." "When Wal-Mart Facility Building Controls notifies your store that a door's open, it means that it's been open over 2 hours."
Because they've have run the numbers of how much that would cost vs just nagging the store to go close the door and then done which ever is most cost efficient.
* https://pentestpartners.com/security-blog/the-fridge-hack-uk...
This includes who talked to whom and what the responses from the manufacturer, supermarkets, and cold store warehouse operators were.
Don't expect to learn what the vulnerabilities themselves are, though. They couldn't get hold of and trace a subset of the device users, so they're holding back what the vulnerabilities are.
Perhaps, you can trigger a hard reset on the device and the firmware returns to default network password?
> Resource Data Management (RDM), the British manufacturer of the refrigeration controllers, disputed the severity of the identified vulnerabilities. The company argued that exploiting them would have required physical access to the hardware.
https://www.aa.com.tr/en/europe/uk-launches-secret-operation...
Which, to me, reads like a secret stored on the hardware. Something like a hard-coded password, storing a single private key on all devices (or a public key with a crazy weak private key), or some other symmetric secret used somewhere in authentication.
The same password used across all units.
The ethical hackers say this would have been achievable as the hardware in question can be purchased cheaply online on sites like eBay."
... TTL serial? CANbus? Some dumb service mode dongle?
I can also understand the purpose of centralised temperature computer control through the local network, but that's not exposed to the internet.
An internet-writeable path seems like inviting disaster.
You might also want control over the temperature to react to electrical prices or grid stability, but that's pretty niche and could likely be 90% scheduled (cool slightly too cold at night, then allow to get slightly warmer (while still food-safe) during day).
I agree though - this should be behind a management VPN, not just on the open internet.
To me this spells out an unprotected private key or secret residing in the device that could be used to break the authentication checks. In which case once secret is out there - no need to posses it? Could the key/secret have been well protected, but the device could be used to sign anything given to it?
from the article ... back door account whose daily password algorithm was seeded from the date
> It was thought the UK had largely escaped such hacks as the British refrigeration industry relies on different devices but there was in fact a potential vulnerability the sector has been rushing to fix.
> In late October, security researchers discovered a major flaw that, if found by hackers, could have led to even worse damage to the UK.
Having worked with industrial equipment like this that often had some kind of network interface retrofitted to a 20+ year old device; I am going to guess it was a manufacture default master password that can be used to remotely manage/reset/reprogram the device and that essentially all deployed units never changed/disabled it.
In fact it may not even have been possible to disable or change it. It is scary the number of appliances, both industrial and consumer, that have "secret god mode" reset passwords and also just happen to work remotely..
Now to defend the manufactures a little bit here these master passwords were originally used because these devices didn't have any kind of network connectivity at all when they were originally design back in the 90s (or perhaps even earlier).
They were there for an engineer to come along, plug in an RS232 or similar and they could use the master password to just fix whatever was wrong. Then over time either networking/internet functionality was added via an additional interface (such as a master control unit all machines were connected to and then that MCU had network/internet access but essentially acted like an unmanaged network switch and just passed on whatever it was sent) or directly to the system in a new model revision.
However nobody deals with the whole master password part and so you end up with a warehouse or supermarket freezer that is now on the internet and if you're lucky "protected" by CG-NAT (usually that is the best case, rarely do you see any real network security in these setups) with essentially a remote login that all use the same password.
Of course this is just a guess from my own experience but if I were a betting man that is where I would put my money.
As for the "security researchers discovered a major flaw" I would also guess this flaw has been known by the manufactures since forever but it was security through obscurity as few people knew of a master password or how to use it.
However most likely aided by LLMs some researchers just dropped in firmware blobs from industrial systems they started looking at after the hacked systems in France in the summer and the LLM did what they do extremely well and find these things.
This set off the alarm bells and the NCSC got involved as a national security concern and thus begins a quiet but rapid call to mitigate the issue before someone else discovers it.
* https://omega-solutions.co.uk
who went actively looking for problems having seen the Danfoss incidents. More detail is in the Pen Test Partners press release (q.v.).
Perhaps I'm being a bit too charitable, but surely if the password is indeed just one single root master password, once it has leaked, its leaked for good and so there's no need to get hold of a device. Or someone could have brute-forced it to begin with...
That's why they put automatic defrost in fridges now. Connecting fridges to the internet, on the other hand....
and these kinds of attacks already happen, consistently. e.g. the Polish Power Grid attacks where they wiped PLCs and disabled network access and observability systems.
and if you're suggesting just killing their wifi to save them then I don't think you understand what or why they installed these systems in the first place.
Which means this stuff is matter of national security in active warfare situation and we know well russia doesn't mind performing acts in other nations that it itself considers acts of war. Which is also fair, russia effectively quietly declared semi-open war with entire western civilization quite some time ago and secret services / governments are all well aware of that. They see us as a mortal danger to their system, which is also fair since it is when talking about their mafia oligarchy.
I wonder why that might be.
> Many important objects and folks were obliterated with say Storm shadow missiles.
Got it.
Its fine I wont put more facts here, thats a meaningless effort for russian trolls / maga folks.
In that time, russia has poisoned british citizens on british soil
Invaded Georgia
invaded Maldova
stoked a war between Armenia and Azerbaijan
Invaded ukraine, twice.
Now, this would be fine from the UK's point of view because those countries are not in the UK sphere of influence anymore. After all the days of thin red lines in crimea are very much behind us.
The issue for the UK is that all these military actions are getting closer and closer to our trading partners. Russia has been very clean on their imperialistic expansion.
Sure spending money supporting ukraine attack moscow is unreasonable, if we just ignore that russia has spend ~100-600k of thier own men in trying to annex the entire of ukraine. And what is next after annexing ukraine? well its poland, hungary, slovakia & romania. All part of the EU trading block, and very much our problem.
Its also not like russsia is being innovative or new here. Its not the first time britian has been sending supplies and support, along with the french to stop the russians doing expansion.
Russia could stop all of this anytime it likes without loosing anything. God knows why they're doing it.
Had he done so I suspect the war would not have continued; the "disputed territories" would still be in an uneasy balance and the world would have moved on despite shooting down a passenger airliner.
Ukraine did not chose war, Putin chose to start and continue the war and is the reason we are where we are today. NATO & the West has been very restrained and not got directly involved in the fighting, Putin though has put N.Korean troops on the front line which is surely a much more serious escalation.
NATO is a defensive organisation, not an offensive one [1].
Ask yourself why the ex-soviet republics are all seeking to join (or have joined) an organsistion that exists for defence amongst democratic countries.
[1] Something Mango-Mussolini was reminded of when he thought NATO would suport him invading Iran.
I guess the reality is someone watching a dashboard scanning the controllers each day and sending emails till they disappear.
Neither is mandatory in the UK, of course. We really ought to bring in some sort of statutory duty to report cybersecurity breaches and near misses, but at the moment we only require that for a relatively narrow set of essential services.
I'm guy, IT guy.
You're more likely to come equipped with a Thermos, toothbrush, extra strength spot cream and be called Dibbley.
Duane Dibbley.
There's probably some Orwellian law in Blighty that gives them the power to command private commercial enterprises like that. That wouldn't be possible in the U.S. and other European countries. In most democratic nations the government can (in peacetime) only strenuously urge companies to update their software, not bully them around like this.
On the other hand perhaps they don't need to each have a connection individually to the internet but could perhaps be fronted by some device that's easier to update and keep safe.
How do we show the most relevant ads to make ends meet with the business?
What if the user needs to know if the fridge temperature when they are at the bahamas with their smartphone via our SmartFridge App[tm]?
1. The OSs, major programs, languages and frameworks are all going to end up more hardened as a result of the bugs that are found and possibility exploited;
2. (Perhaps more importantly) The systems that companies are governments absolutely rely upon are going to have to deal with security and reliability in a way that they largely never have before. Such things were and still are treated as an unfunded mandate for which nobody is rewarded for fixing these things and nobody is punished by ignoring them. That era, I believe, is over. That's not such a bad thing.
Victim blaming. It's very easy to overpower an old lady in public, but it's wrong to do and it's a criminal act
I'm not saying that security shouldn't be improved, it should. But the fact there's a spree of opportunistic criminals because crime became easier for them to achieve doesn't make it not a crime
It's no different to, say, (correctly) blaming a car company if their car under certain conditions can cause a steering lock or spontaneously catch fire when it's too hot outside. Blaming the victim would be blaming the individual drivers.
If a car maker puts the strongest locks on the door and ignition, and the owner gets out leaving the car running and the door open, and you slip in and steal the car, that is also a crime
Stealing cars is a crime no matter how easy it is to do. Don't give me this "well it was easy for me so it's the car makers who are the real criminals" as you drive away having stolen my car.
The people breaking into fridges, regardless of the strength of the security, are criminals, and what they are doing is a cyberattack. It doesn't matter how easy it was for them or how negligent the fridge vendor was.
The corporation bought in from reputable vendor ... in your analogy they bought a car (fridge thermostats) from a car company (fridge thermostat company) that blew up on them
Maybe some people from the ice industry took the proliferation of home refrigerators badly.
Quite a few owners and operators didn't know these controllers were directly exposed to the internet. Smaller operators often had them on cellular modems and didn't know they were online at all. Some larger operators assumed they weren't exposed. They were wrong.
Then you have to work out who owns each one. There's the retailer and the refrigeration contractor. In many cases these were franchise stores. Seeing a device on the internet doesn't tell you whose it is. Geolocating the IP means very little when it's behind a cellular modem.
Then you have to contact them. "Your refrigeration controller is exposed to the internet and might get hacked, can you disconnect it?" is an odd phone call to get.
Then someone has to physically find it. Supermarkets don't have IT staff on site, and you don't send IT into a refrigeration plant room. That's travel, logistics and organisation, site by site.
And you have to know that disconnecting it is safe. If that controller runs the refrigeration for the whole store, how do you know pulling the Ethernet cable won't cause a serious problem? It all needed risk assessing.
This was a significant amount of work across a lot of organisations. It wasn't putting your trousers on before leaving the house.
I guess the justification is that the public certainly has an interest in supermarket fridges not becoming space heaters.
And the damage inflicted in some European countries was limited. There wasn't any widespread food shortage or anything.
It was simply a result of negligence on the part of the manufacturer and supermarkets. Their claim that they make so little markup on food stuffs that they can't afford decent IT security sounds silly to me.
[1] Food security:
https://en.wikipedia.org/wiki/Food_security
Security Researchers announce that BigCo leaves a key to the front door under the door mat!
BigCo denies that they do that, announces that security is their middle name and that they have corrected the problem.
Security Researchers announce that BigCo now leaves the key under the flowerpot.
BigCo announces that security is of utmost importance to them and have corrected the problem.
Security Researchers announce that BigCo now puts the key in an envelope labelled "Opening This Envelope is a Federal Crime! And there is no key inside!"
Security Researchers roll eyes.
BigCo sues Security Researchers, accuses them of terroristic acts and says that G*d wants them to make ever increasing quarterly profits and that they can't make money if they don't move fast and break things, and that they definitely did not use an LLM to create their security architecture.
Directly sending monitoring/metrics? Maybe. Directly controllable? No! Absolutely not! At minimum, require a VPN tunnel to the fridges' network. They same should be true for anything that doesn't have to provide network services to the public.
Same goes for all IoT hardware (the S in IoT stands for security)
Is a vendor selling you a "smart" device, where the "smart" part happens only if you hook it up to the internet, and can be unilaterally changed at any time by the vendor because all the logic's on their end, not yours? Tell them to fuck off and put the "smart" on the device itself or you're not buying it, and while you're telling them that, tell them to fuck off with their subscription-based or subscription-alike business model, it's blatantly being done in the vendor's interest and not yours.
They may be morons, but the HYS crowd have entirely the right idea.
For the first 150 years thermodynamics and refrigeration didn't need the internet, so why now?
Frankly I've no sympathy for those who have to have to manage this problem now. As I see it, this is yet another instance of using technology for it's own sake—just because it's now doable.
Connecting them up to an integrated logistics system opens up big savings. You can start cooling only as a delivery is on the way and at temperature when it arrives.
Britain has historically had some of the best run supermarkets in the world. This is a way to pass savings on to customers.
It's one reason why fresh produce is generally cheaper in the UK than the USA even though the cost of energy is much higher in the UK.
A local thermometer won't know that lory with an exact number of pallets is held up in traffic or what its ETA is.
An integrated logistics system can know where every lory is, what's on it and where it needs to go.
It can control the fridges for drop off along the way and can deal with changes in delivery schedule or on demand re-routing.
Imagine customer demand changes in one particular shop. In another one fridge isn't reaching temperature, re-routing a delivery will wastage whilst a servicing team is sent to fix it.
Without that the fridge will lag in getting down to the correct temperature for arrival.
I guess you could have drivers pull over and call into the shop with their location but I'm sure you can see that's open to potentially costly mistakes.
I get what you're saying but we're talking about industrial scale logistics and cold chain management. 150 years ago I wouldn't have been able to get whatever fruit I desired year round from the supermarket and honestly I like that I'm able to now and consider this real tangible progress.
Also with the manual method it needs to be accurately written down, accurately entered, and then analysed to see if there are any problems building up, not just obvious failures. Having estate wide real time data can make a big difference.
There can also be a lot of data collected to predict problems. If you detect one unit using more energy to hold temperature then perhaps a component is failing and you can get it serviced before it breaks. They can also monitor efficiency, perhaps half of the fridges get the morning sun and struggle to maintain temperature so putting window blinds may help.
Question. What do you think of remote monitoring of computer servers, surely someone can just walk around the data centre, and your customer offices, with a pen & paper checking for warning lights. For example HPE SmartArray predictive failures of hard drives, who needs that!!
As I implied elsewhere, it's a problem of their own making.
Sorry, they were not. How could they be when TCP/IP still had to be invented? Moreover, many used HW that was physically incompatible with the low levels of the OSI model (the physical layers couldn't even connect).
A well designed modern network architecture is perfectly capable of providing a much higher level of security than any remotely-accessible system from the 80s or 90s.
The problem is the lack of that well-designed network architecture, not that modern technology somehow makes it inherently impossible to secure.
Not really.
As a kid, I had a saturday job in a supermarket...... in the pub after work, we'd all be drinking away with our boss who would typically be "doing the freezer checks" and writing random temperatures in the log book at the pub table.
Happy days :-P
"Who the hell wants to go peer at fridges all day and has enough incentive to not just sack it off" is my immediate response to those suggesting we can take a step to full employment by not having these things do it automatically.
In the old days (say 20 years ago) energy was much cheaper at night, so to save costs, one of the things you could do is drop the setpoint of the freezers 10 degrees to extra cool the freezers, then set it back to normal during the day. This would save not insignificant amounts of cash. So long as the stuff didnt unfreeze, it doesn't affect the quality of the food (within reason, there are caveats, that someone more knowledgable can chip in with.)
Nowadays the the cost of energy is less predictable, but avoiding significant draw between 7-10am and 16:30 and 20:00 can half your energy costs. (https://agileprices.co.uk/?region=C this is customer variable pricing, but peak time is 2x the cost of offpeak, sometimes more.)
It's been known almost since the inception of the internet that putting systems online that used the standard internet made them highly vulnerable, so it's why critical systems should not be put online but use proprietary methods/protocols that cannot be easily hacked, especially so if used with specialized hardware built for the purpose.
It's clear to me this was either cheapskate design, or fashion, or done for convienience—i.e. check the status with a laptop or one's phone.
We've seen this security shortsightness previously, for example axing the job of the dam keeper who'd always be present to open and close dam sluice gates just because someone could monitor the dam online from afar—just to save a few bucks.
It's not as if these vulnerabilities weren't foreseen, they were. Now these dumb (and greedy) planners are paying the penalty.
Putting fridges behind a VPN means an investment and installation in each physical store. Now, some of these super markets are big chains, others are small owners with a franchise model.
3,000 stores - 24/7 fridge checker - £100mil a year in cost per large supermarket to check to fridge temperature?
what happened to the hacker spirit that would see that as the basis for a business selling a subscription for chopped fruit in a bag you squeeze at home?
For example: So the service company can monitor temperature and get fault alerts and error codes remotely, shops can change temperature (for different foods) and internal lighting (when shop opening times change) without having to visit each unit on their hands and knees.
I'm sure, buried in the details, is also uk.gov's weird and wonderful health and safety laws because of food standards, that somehow requires remote access to temperature control (e.g. 'all refrigerated food needs to be at -10.25C' - one click and you're done).
I don't think it is unrealistic or unreasonable to dictate that if my pack of frozen chicken is unfrozen for 48 hours its not longer healthy or safe
That's your problem right there; if you only know how to use a hammer, then all your problems look like nails.
It'd be enough to have a bog standard segregated network with a basic security elevation mechanism, together with some reasonable amount of logging and monitoring. The sort of configuration that tech companies use to protect their database server.
Data diodes make sense in situations where you need security clearance to even be in the same room as someone who's accessing the high side - they're expensive and painful to use, and nothing a supermarket does is ever likely to justify it.
Want to know the temperature, remotely, for whatever reason? Put a 4-dollar sensor in it, and have it report to a read-only homeassistant-style setup. preferably behind a proper firewall.
https://news.ycombinator.com/item?id=50335615
https://news.ycombinator.com/item?id=50159616
And going with the 'one peanut less saves millions', at the corporate level, a bean counter probably figured out being able to remotely change the temperature of every fridge consistently across hundreds of supermarkets by a tenth of degree saves millions in power.
Absolutely everything can be achieved by clerks if that wasn't unpopular.
Pardon the Reddit image link, but: https://i.redd.it/17b6m5a31q9c1.jpeg
Walmart internal propaganda stickers on the back of house refrigeration. "For every minute the door's open, it costs your store 55 cents." "When Wal-Mart Facility Building Controls notifies your store that a door's open, it means that it's been open over 2 hours."