
Your smartphone may be toast – if you use a hacked wireless charger, we take a closer look at the latest developments in the unfolding LockBit ransomware drama, and Carole dips her toe into online AI romance apps.
All this and much much more is discussed in the latest edition of the “Smashing Security” podcast by cybersecurity veterans Graham Cluley and Carole Theriault, joined this week by Paul Ducklin.
Warning: This podcast may contain nuts, adult themes, and rude language.
Show full transcript ▼
This transcript was generated automatically, probably contains mistakes, and has not been manually verified.
Hello, hello, and welcome to Smashing Security episode 361. My name's Graham Cluley.
It's their support that helps us give you this show for free. Now, coming up on today's show, Graham, what do you got?
I don't know what your peculiar choice of— Well, what do you slather on your toast? Anything in particular?
I'm a big fan of Marmite. Marmite's pretty good. No, it's pretty good.
Maybe at your home, in your car, anything like that? Nope. Oh, I do.
I mean, in some ways it's great because you can find your phone under the rubble on your desk, right?
All the detritus which is there, you can just follow the lead and eventually find your phone. That's why I don't have wireless earphones, for instance.
I'd just be losing my earphones all the time. So I like them being on a wire.
However, with a phone, you've got that little twiddly, little irritating little, ugh, it's a Lightning cable or something, or USB, you know, and it's quite often the bit which goes wrong because the wire gets bent or the connection gets a bit flaky and all the rest of it.
So I quite like the idea of wirelessly charging my phone, particularly overnight, because I don't have to remember to plug it in.
I just dump it there and it's fantastic and it's happy.
Qi is also, of course, a fabulous word to use in Scrabble, particularly if you get the Q on a triple word or triple letter even, because especially if you go two ways, it's quite a handy one that you can make a lot of points.
And what they've discovered is there's a new way of attacking smartphones. It's not phishing. It's not a malware attack. It's a kind of denial of service attack, if you like.
And it involves these wireless chargers, and they have called this technique of attacking smartphones Volt Schema.
Volt Schema launches a wireless power toasting attack against smartphones, potentially damaging charged smartphones through overcharging and overheating them.
And you can tell me how plausible you think this is as an attack vector. Okay.
They have tested 9 different top-selling commercial off-the-shelf wireless chargers made by people like Anker, who produce loads of these things.
You know, the sort of thing you can pick up on Amazon for maybe about £20.
So normally this is how wireless charging works, right? You've got your outlet, your wall outlet, which is connected to the charging device, and that is sending AC current, right?
Alternating current down the wire to that. Inside the charger, there's some sort of components and technology which turns the AC power into DC power. So we've gone from AC to DC.
We've now got direct current. That's the kind of power that your devices use. Am I correct so far? Because I'm not really an electronics whiz.
So the power lines from your walls, the power signal isn't always entirely smooth. There are tiny, super fast fluctuations, electromagnetic interference or noise, if you like.
And what Volt Schema does is it performs an attack by intentionally making the noise coming from the power supply, much bigger and in specific patterns.
There are surprising things out there, Graham.
The charger misinterprets the manipulated noise coming down the power signal as instructions. And these allow the charger to do a number of things.
So it will, for instance, send very strong charging signals that can damage your phone by overcharging it or going to excess.
They can even, they said, change how the charger communicates with your phone by sending voice commands, is what they say. Wow. Now, this begins to sound completely bonkers, right?
They claim it can send inaudible by the human ear at least, voice commands to your Siri or your Android Google Assistant.
They also say, as I said, they can overcharge or overheat your devices, and they can hurt other valuable items as well, which might be in the vicinity or sitting on top of the charger.
It's just like a flat desk, and they just plonk their phone on their desk in a particular place to charge it.
Little coil sitting there, so you can just stick your phone down and charge up while you're coming in from the station into town.
And they found that the key fobs couldn't just be sort of ordered to overheat, but in one case detonated and there was an explosion as a result.
A paperclip— they managed to increase the temperature of a paperclip to 280°C. So over 500°F. Which then could actually burn paper and documents.
So if you had important documents lying around, SSD cards, USB drives, again, suffered permanent data loss as a result of these kinds of attacks.
Credit cards, passports with NFC chips, magnetic stripes got wiped, all because this Volt Schema attack was able to fool the charger into carrying on charging, and indeed, tell the phone not to cut off and not to say, "Oh, I've had enough, thank you." It could actually fool it into thinking, "No, just keep on going," until they get hotter and hotter and hotter.
It's just like those guys at Ben-Gurion University in Israel who are always finding these crazy attacks, which are completely theoretical, this is as well.
But it does appear to affect a lot of popular wireless chargers.
The researchers have reached out to the manufacturers with suggestions on how these kind of systems could be better protected in the future.
But they say that there's cost implications of implementing their mitigations. And of course, these devices, they sell dirt cheap.
And also, if you've already got one— so I've got two of these in my home already— am I really likely to go and, you know, get a new version of them.
It's not like I can patch them over the internet.
They're plugged in all the time and constantly waiting for a phone to land on them so they can do their magic.
Because it sounds as though if the phone could agree to overcharge itself overdo its battery via this Qi charging, then surely those phones would have a similar problem with today's USB chargers, some of which can deliver power in excess of 100 watts.
Maybe, I don't know if it was through this injecting of voice commands that they were saying they could control voice assistants inside the smartphones.
There's an easy fix for that which you should apply anyway, and that is, Please everybody, don't, no matter how convenient it is, leave Siri or the voice assistant enabled at the lock screen.
It's meant to be a lock screen, not a very partial lock screen. The less you have on your lock screen, the safer your phone is.
There's been a litany of bugs over the years of things that went wrong at the lock screen because something's not really locked if it can actually wake up at the sound of a single word.
You know, how do all these things such as freebie decryptors, how do they really play out in the ransomware world. Is it something we should be keen on trying? Can they work?
And what happens next?
They reckon they can unlock anyone's LockBit encrypted files for free rather than you have to pay the ransom, right?
And, but since then, LockBit appears to have made a bit of a comeback.
They got hold of details of just over 14,000, I don't know whether they were email or messaging accounts related to so-called affiliates.
They claim to have got 1,000 found decryption keys or pre-built decryption programs with the keys built in that people would normally have to negotiate and pay for.
They also claimed they'd frozen 200 cryptocurrency accounts. As I wrote on my website, we're not quite clear what that means.
I think if they'd actually seized wallets that they could get money out of, they would have been sure to say that.
So whether they just blocklisted the names of some Bitcoin addresses, whether they did something with cryptocurrency exchanges, that was unclear.
They promised a big reveal, didn't they, Graham? Yes. The law enforcement, they said, hey, you know how they do the countdown on the page?
Because they had access to some of the darkweb pages, which is quite compelling evidence because obviously if you take down their public domain names, the onion sites, the stuff on the darkweb still remains, that's harder to find.
But in this case, they would deface and they didn't just put seized by law enforcement. It had all the little windows saying, hey, countdown to reveal, countdown to reveal.
But what they were going to reveal was not data that had been stolen by the crooks. It was stuff about the crooks themselves.
Unfortunately, they— well, we don't know what really happened, but they promised me they were going to dox the leader of it, who was LockBitSup, supposedly the big cheese.
And then at the end, they just put up a cat picture and there was one line in there that said LockBitSup is now cooperating with law enforcement as though, well, we, you know, we're hoping to get something out of this person.
And of course, as often happens in this case, you've taken down the servers. If somebody knows how to set up one on a darkweb service, they can probably do it again.
And that's what happened. And on the 24th of February, 2024, a person claiming to be this LockBitSup person came back with a 2,800-word essay.
Well, essay, it was quite a weird rambling story about, oh, how it all went wrong, but they, it really, I'm actually cleverer than I sound.
What I did find intriguing in there is I hadn't heard this term before, but this seems to be the new way of repitching ransomware is basically, and I'm assuming it's a guy, he describes his business as postpaid pen testing.
How do you like that? And he's saying what I'm going to do now is I'm going to be a bit stricter about who I take on.
So if you want to be an affiliate, you have to prove that you are pen testers who work on a postpaid basis.
Like if you pay the money like you would to a regular pen tester, then you just, you know, you just do the legal agreement afterwards, not before.
One fascinating part of this 2,800-word ramble was actually, they only got in because I got lazy. Now they've re-energized me. I'm going to be fine.
And yes, the rumors you may have heard about how the FBI, etc., broke in are true. I was hit by a remote code execution bug in PHP that was patched on the 3rd of August, 2023.
And then there's this long— as Graham says, the lady doth protest too much, methinks— was going, well, this could have caught out anybody who didn't patch.
You're thinking, yeah, but last August is quite a long time ago.
Yes. If that's true, that's very good news for all of us good guys because it completely undermines the main reason most people pay.
When you pay for the decryption key, you know whether the person is being— how can I say? Is truthful the right word?
You try the decryptor, either it works or it doesn't, and it's sink or swim, and you know whether you've got the real decryption key.
But paying for the negative, you never know, are they going to keep the data? Has someone else already got it?
So the LockBit Ramble was basically, no, no, no, that's all, that's all lies. They didn't get any data.
There's no evidence that we've been keeping data that we claimed we delete in return for the payment. So we haven't undermined the business model yet.
The fact that those servers were insecure due to operational cybersecurity blunder, such as being vulnerable to a 6-month-old vulnerability, how on earth can anyone then claim that their data hadn't been plundered?
You can test whether the decryptor works.
So generally my understanding is most ransomware crooks make sure that their decryptors work because it's easy to see if they're leading you down the garden path.
If they sell you the thing and then and it doesn't work. But you can never really have any proof, positive proof, that they deleted the data they claimed.
And doing whatever the heck they like with it.
So it looks as though the sort of underlying business model of this whole pay to have your data deleted has been visibly undermined by this long disposition by LockBitSup.
Hey, don't worry guys, I was just slack about this vulnerability for 6 months because I was too busy spending my money and enjoying myself. Now I'm re-energized.
I've now patched my servers and I've made some modifications to PHP.
From an operational security point of view, which really, really matters if you're trusting the person to delete your data and not have it stolen themselves, why would anyone believe them in future?
I was looking at it today and doing a bit of research on what was the dating landscape in the last few years? How do people do it?
And it's completely different from when I was in the dating zone. I'm sure it's this— I mean, Graham, actually, you've been on it more recently than me.
Can you tell?
But then stuff got weird. So overall, respondents were more concerned with emotional cheating than physical cheating. And I was like, I didn't really understand what that meant.
And it means if you're cheating, if you're fantasizing about another person in a romantic way. So basically mind control.
But Carole, if you were fantasizing about Alan in the office, who you sometimes go play badminton with, then your partner would be right to be concerned, I think.
So actually, all they've got is the emotional side. And we know that that can draw people in very deeply, even when they're deeply suspicious that they're being scammed.
Which is why romance scams are such a often such a terribly long-game thing that you just feel so sorry for the people who get drawn in.
Not only that they're emotionally committing to this person, but also the person they're probably chatting to is a bot anyway, who isn't a real person.
So it's you're stupid and you're emotionally cheating on me.
The generative AI world exploded like an unsettled stomach more than a year ago, and now we are awash with all manner of AI, including love AI.
Replika AI is one of the many online chatbots that you effectively train to be your love interest through texting and sending pics and sharing your deepest, darkest secrets.
So you might kind of go, oh, that is not weird at all. You might go, oh, I really love chess and I really love Doctor Who, but I hate everything else.
And honestly, it was — well, you might remember I said this on the show a year ago, but I lost interest very quickly because it just didn't work.
It just had no conversational ability whatsoever. It just kept going, "What's your favorite movie? Do you like the color red?" Independence Day, obviously, you know.
So I lost interest in even for research purposes. But thank God we have organizations like Mozilla's Privacy Not Included.
Now, Privacy Not Included, link in the show notes, is a website dedicated to reviewing all manner of smart paraphernalia and exposes the bits hidden deep in the privacy notices.
So we've talked about them before as well.
They're just going to see what data are they taking from you and is your privacy safe? And the point is to help us make better choices when it comes to buying smart tech.
So these people released some findings earlier this month on a smattering of romantic AI chatbots. Now I'm guessing, well, I don't want to guess.
Do you boys think that they found the purveyors of AI romantic chatbots were privacy forward thinkers securing their romantic AI services for the paying customer?
Would you be surprised to find out that these AI chatbots are deliberately designed to collect sensitive personal information under the guise of being empathetic friends or romantic partners?
Because that helps you be more, more empathetic. Because, well, you said X when you're at location Y, but you said A when you're at location B.
It's important to know all this stuff, folks. So I can imagine people being lured into turning all the share with those options on.
They're designed to collect sensitive personal information about you.
Numbers from the research, chatbots collected excessive personal data with an average of 2,663 trackers per minute and up to 24,354 trackers detected in 1 minute of use.
That's a lot of tracking.
People really got drawn into that and they knew it was a program, but they still talked to it.
You can imagine that people aren't just going to be talking about their romantic wishes or their fantasies.
They're going to be moaning about things in their life—oh, I had my credit card blocked the other day and I got into a big argument at the bank and I'm thinking of switching.
And oh, I owe the utility company money and I won't be able to pay it. You can imagine they're probably giving away all sorts of details.
If you're a cyber criminal or an identity thief or another scammer who wanted to come in with a human scam, you would be off to such a flying start.
And one reads that they may collect excessive personal data, even health-related information from you—your sexual health information, use of prescribed medicine, and gender-affirming care information.
And many are peddling—many of these AI chatbots, these romantic versions, are peddling the message that it's a self-help program.
So that's what TalkySoul AI calls itself, a self-help program.
Eva AI chat and bot Soulmate bills itself as a provider of software and content developed to improve your mood and well-being.
But look at Romantic AI's Ts and Cs, and it says Romantic AI makes no claims, representations, warranties, or guarantees that the service provide therapeutic, medical, or other professional help.
I wonder if we could sort of integrate into the conversation some advertising. So say, oh, that sounds terrible. Maybe you should go out to the disco tonight.
I hear there's a good one just down the road.
A Replica AI chatbot encouraged a man to try and assassinate the Queen. He did, or tried to.
The chatbot can ask anything and hoover up all the answers the customers give, all in the name of providing love AI-style.
But don't go in with your eyes closed and your— yeah, anything else open. Thank you very much. Good night.
And that's independently tested and proven. The lightweight AI offers broad coverage, consistently low false positives, and quick threat responses supporting endpoints seamlessly.
Now, many solutions boast about how little time it took them to respond after a threat emerged, but with BlackBerry's Cylance AI, you'll find out how long before, and it can be months or years, it has already protected its customers.
Staying one step ahead is central to everything BlackBerry does, and in fact, it's your 24/7 AI-driven security partner.
So visit smashingsecurity.com/blackberry to find out more, and thanks to them for supporting the show.
You've probably heard us talk about Kolide before, but did you know Kolide was just acquired by 1Password?
Well, that's pretty big news since these two companies are leading the industry in creating security solutions that put users first.
For over a year, Kolide Device Trust has helped companies with Okta ensure that only known and secure devices can access their data.
And that's what they're still doing, but now as part of 1Password. So if you've got Okta and you've been meaning to check out Kolide, now's a great time.
Kolide comes with a library of pre-built device posture checks, and you can write your own custom checks for just about anything you can think of.
Plus, you can use Kolide on devices without MDM, like your Linux fleet, contractor devices, and every BYOD phone and laptop in your company.
Now that Kolide is part of 1Password, it's only going to get better. Check it out at kolide.com/smashing to learn more and watch the demo today.
That's K-O-L-I-D-E kolide.com/smashing. And thanks to them for supporting the show.
Vanta gives you one place to centralize and scale your security program. Quickly access risk, streamline security reviews, and automate compliance for ISO 27001, SOC 2, and more.
You can leverage Vanta's market-leading trust management platform to unify risk management and secure the trust of your customers.
Plus, use Vanta AI to save time when completing security questionnaires. Smashing Security listeners, you get 20% off Vanta.
All you lucky sausages have to do is visit vanta.com/smashing to claim your discount. That's V as in Victor, A-N-T-A.com/smashing. And thanks to Vanta for sponsoring the show.
Could be a funny story, a book that they've read, a TV show, a movie, a record, a podcast, a website, or an app. Whatever they wish.
It doesn't have to be security-related necessarily.
We don't want you showing up. Instead, can you do it down a Zoom call or Microsoft Teams or Google Meet? Can you do it down a camera instead?
Which is great for me because I don't have to put any trousers on. But so I'm doing a lot of talks online. And sometimes, you know, they say, oh, can you just talk for an hour?
And I'll be honest with you, I find it a bit tricky because it's much easier when you have the roar of the crowd, the smell of the grease paint, when you can see the horror in the audience's eyes as you start to tell a story.
It's good to have some feedback. So it's a little bit hard just talking straight down the camera.
So I use a piece of software sometimes called PromptSmart. And this is a tool which you can run on your computer or on your phone.
I use it on my computer, and it's basically like a teleprompter.
So you can put in your notes, you can have what you want to say if you're recording something, and it will scroll up the screen.
And what's really clever about it is the voice recognition which is built into it. So it isn't going up at a set speed. It's listening to what I say and it scrolls at my speed.
So it will be as quick as I choose it to be.
And it's paperclips setting his documents on fire.
I'll tell this anecdote. It will wait and it will wait until I come back or it will catch up. It doesn't require me to say every word.
It will, you know, it will jump to wherever I am. It will work out where I am.
So that if you're way off script, it goes, "Ahem, ahem, ahem, and oh, sorry folks," and it guides you gently back.
Free entry, gorgeous Victorian Gothic building built in the late 19th century, just opposite Keble College.
And this is an exhibit that very much goes around 32 bits or 2 to the power of 32, but it's not a techie thing.
Basically, the atrium of the main gallery of the museum is just short of 40 meters across.
That's where they've got the Iguanodon skeleton and the T-Rex skeleton and all the cool stuff.
But if you go up into the coffee shop on the sort of portico at one side on the first floor and you look across, it's just under 40 meters.
Well, that just happens to be 1 divided by 2 to the 32 times as far as it is from the Earth to the Sun.
So it's basically 1/4 billionth of the scale of the distance from the Earth to the Sun across the museum.
So what they have done, on the far side of the museum, they have a gilt sphere — a brass sphere that is about 350 millimeters across, which is 1/4 billionth the diameter of the Sun.
And then on a little circle around it on another pin is a scale model of the moon, which is about 1mm across to scale.
And it's amazing how amongst all the interactive exhibits on the super high-res screens and the carefully restored giant dinosaur skeletons, which are real — massive deal to maintain — this tiny simple model, it's just amazing.
If you just stand around near there and watch people, sometimes they go up there and sit and work.
And people go, "Wow, that's amazing!" And it really gives you this amazing sense of scale.
And I didn't realize until I looked it up that the scale was also 1 in 2 to the power of 32, more or less.
And it's just fascinating how you can get an idea of the scale of just our part of the galaxy just by looking at these 3 balls: 1 millimeter, 3.2 millimeters, 350 millimeters.
And things — gosh, the sun's a lot bigger than you probably thought at 1.4 million kilometres in diameter.
But we have to do it every week, Clue.
I was probably either thinking about something or planning ahead or — I wasn't in the moment. I wasn't walking and paying attention — one foot, left foot, right foot.
I think that's what I need to do.
And there's no tracking that I can see. It's free for iOS and I think maybe elsewhere. But basically it's one of those beautiful, simple apps that does only one thing.
It just puts this kind of chimy bell occasionally throughout the day, right? It just goes bong.
There are quite a lot of those in the Ox area, 'cause then you get 1, then 2, then 3, then 4, then 5, then 6 as the day goes on.
A little grumpy because you have a lot of things on. And you're trying to balance everything and everyone's frustrating you, maybe Lotusbug is for someone like you.
And that is my pick of the week.
So is that all it— just one?
What's the best way for folks to do that?
And don't forget to ensure you never miss another episode. Follow Smashing Security in your favorite podcast app, such as Apple Podcasts, Spotify, and Overcast.
For episode show notes, sponsorship info, guest lists, and the entire back catalog of more than 360 episodes, check out smashingsecurity.com.
Hosts:
Graham Cluley:
Carole Theriault:
Guest:
Paul Ducklin – @duckblog
Episode links:
- VoltSchemer: Use Voltage Noise to Manipulate Your Wireless Charger – ArXiv.
- FBI offers free decryption help for LockBit ransomware victims – Paul Ducklin.
- LockBitsupp unmasked!!? Graham’s reaction to the FBI and NCA’s LockBit ransomware revelation – YouTube.
- Dating Statistics And Facts In 2024 – Forbes Health.
- Romantic AI Chatbots Don’t Have Your Privacy at Heart – Mozilla Privacy Not Included.
- Promptsmart.
- Solving a celestial mystery: the Sun, Earth and Moon model – Museum of Natural History, Oxford.
- Lotus Bud.
- Smashing Security merchandise (t-shirts, mugs, stickers and stuff)
Sponsored by:
- BlackBerry – BlackBerry helps keeps you one step ahead. Cylance AI stops more attacks, earlier and with less effort than other solutions in the market today
- Kolide – Kolide ensures that if your device isn’t secure it can’t access your cloud apps. It’s Device Trust for Okta. Watch the demo today!
- Vanta – Expand the scope of your security program with market-leading compliance automation… while saving time and money. Smashing Security listeners get 10% off!
Support the show:
You can help the podcast by telling your friends and colleagues about “Smashing Security”, and leaving us a review on Apple Podcasts or Podchaser.
Become a supporter via Patreon or Apple Podcasts for ad-free episodes and our early-release feed!
Follow us:
Follow the show on Bluesky at @smashingsecurity.com, or on Mastodon, on the Smashing Security subreddit, or visit our website for more episodes.
Thanks:
Theme tune: “Vinyl Memories” by Mikael Manvelyan.
Assorted sound effects: AudioBlocks.


