FBI taps cell phone mic as eavesdropping tool (2006) | Hacker News Reader

FBI taps cell phone mic as eavesdropping tool (2006)

cnet.com

55 pointsmmastrac93 comments

I just assume everything I do or say is recorded and stored forever on some redundant storage and can be easily searched if I'm ever notable enough to have influence. That shouldn't influence what you do or say though, because if a system like that is corrupt enough to destroy a person for their private conversations, it's not worth participating in or supporting in the first place.

> can be easily searched if I'm ever notable enough to have influence

Simply by voting, or working for a certain company, or or striking or joining a union, or seeking an abortion may very well be notable enough for them to label you a certain way so they can pin something on you - "Acting against state interests" or some such.

> Simply by voting

Please show me a instance where a judge approved a warrant to investigate the crime of voting.

The whole problem with this shit is they don't always need a warrant.

If you are a non-us citizen or you're deemed "a terrorist" they can do whatever.

This complicates further with crap like the US spying on say Swedish residents and then forwarding information to Sweden. Technically; Sweden isn't spying on it's citizens, the US is. But it's the same result.

Or they can just buy the data they want from whatever surveillance capitalism megacorp and not even have to bother collaborating with international orgs, regardless of citizenship/national origin/whatever.

https://www.texastribune.org/2022/05/11/crystal-mason-illega...

I guess that’s a good point. Not everyone can vote so there absolutely should be investigations into people bypassing the law.

I assumed “legal” voting from the original comment but I guess its not all legal.

> so there absolutely should be investigations into people bypassing the law

No, there really shouldn't have been an investigation. This was evil, wasteful, and above all, truly fucked up.

She didn't "steal" a vote from anyone. She cast a provisional ballot, after asking permission. There doesn't need to be an investigation- just check the facts, and deal with her provisional ballot as necessary.

At my college, we had a guy walk across the stage at the graduation ceremony for the school of engineering. But he wasn't in the school of engineering; he just found there ceremony to be at a more convenient time, and assumed it was all the same. They didn't call his name, and nothing came of it. Because it was an administrative issue. What they didn't do, was throw him in jail for trying to steal an engineering degree. Because there was zero chance of that happening, and he wasn't even trying to do that.

That's not what is being debated here. We are debating whether judges regularly sign off on warrants for legally voting or voting for the wrong candidate.

Nobody said "regularly", or "legally" or "wrong candidate" until you, just now. Feels like goalpost moving. But if you think my example was pointless, feel free to ignore it. I answered the question that was asked, and now I'm done.

> Simply by voting

This means voting illegally?

Nope, it means "simply by voting".

Whether the vote is valid or not, whether it should be tossed out or kept, is a separate issue that really should be determined asynchronously, and carry no penalty. Especially if, as in this case, she literally asked the authorities if it was okay, and her ballot was only ever provisional in the first place.

If some votes are illegal how could investigating illegal votes be unlawful in general?

I didn't say it was unlawful to investigate. I said it was unwarranted and evil.

If there's a list of people who are allowed to vote, what's the fraud here? What exactly are they getting away with? It's like the crime of seeing a "Sorry, you must be logged in" page on the web. Oh noes! That page is only for certain people, and you're not one of them, and you tried to see it and failed! Somebody call the cops.

I was talking in the general sense. Not about 1 specific case. Thats what you focused on.

I'm with you for the most part, but your examples don't seem to represent the type of citizens that the IC has been targeting for at least the last decade.

Still, we should never allow the government to do to our opponents what we wouldn't want done to our allies.

So called white supremacists.

> type of citizens that the IC has been targeting for at least the last decade.

The really crazy part is that none of us has any clue what type of citizens the IC has been targeting.

I'd be willing to bet it's all of us.

Yes, I just found out, at least Samsung Android, my phone does OCR on every picture ond does store it somehow in a DB. When I search for something on my phone, pictures with a part of the search string pop up instantly. Even bad pictures from a monitor, every word is indexed.

Of course .. I'm pretty sure, everything goes to Google / Samsung, you know, to help to make a better service.

I wouldn't be surprised if your phone was sending a copy of every picture you take to third parties too. Roku, for example, takes multiple screenshots every second of whatever happens to be on the screen and sends them back to Roku who analyzes them to guess at what you're into and to identify the content.

I one day discovered that my phone's Samsung keyboard (the default one that came with the device out of the box) was logging every key press and sending it to a third party whose privacy policy said that they used it for "market research" and to guess at things like the intelligence and education level of the user. I immediately installed a new keyboard. I went with AnySoftKeyboard and so far it's been great, although, understandably, the autocorrect isn't anything close to what it used to be.

Yes I know. The keyboard was the first thing I changed.

Is that Google Photos? It does it by default if you back up images there, and can be disabled.

Many people end up using Google Photos just because they have an android phone, it was never anything explicitly intentional. And it definitely does more than OCR, a buddy wanted to show me some pics from their weekend and typed "kyak" into the search bar and voila, there were their photos of them in a kyak.

I don't use Google Photos, since it is, it was, broken without internet access. I block Google Photos with NoRootFirewall. Some years ago so it stopped to show me even local pictures. But yes, since it's not possible to deinstall, it is still there ...

> That shouldn't influence what you do or say though

It shouldn't, but I'm willing to bet it does. Even if you don't notice it. In the best(?) case it influences a small number of people you talk to, but that influences your conversations with them so indirectly influences you.

But I'd also argue that such a system would also be equally powerful enough to convincingly fabricate evidence against you. They could still blackmail you, by identifying things you might be embarrassed about, but the ability to fabricate convincing evidence is still highly problematic.

I do often wonder if the next big boom in "ad" technology will be homomorphic encryption. Where the data can be mined for the useful things like data to train ML models or target populations of people, but targeting individuals would be substantially harder.

> I just assume everything I do or say is recorded and stored forever on some redundant storage and can be easily searched if I'm ever notable enough to have influence.

That's the real problem with constant surveillance. Those in power can use it to find and silence anyone who they feel is a threat to them, or even anyone who they think might one day become a threat. I'm not sure that freedom and democracy can exist under conditions like that.

And yet, how the hell does so much serious crime still remain in the world?

I did a fun experiment with my iPhone 12 recently.

I put it in my living room in a plainly accessible area not obstructed by much. I said in a normal speaking voice "hey siri" and waited to see if it activated.

Then I stepped farther back, and farther back, and I was shocked and honestly disturbed at the distance it could clearly pick up my voice saying "hey siri." It was nearly 2 room lengths away, speaking at a normal volume, and it was very easily picking me up.

FWIW the assistant activation training is far more sensitive than everything else. It's designed to pick up the activation words in all sorts of situations, even if it can't understand what you say afterwards.

That's because it's using the phones processors, no? In a hypothetical fbi is recording scenario they might have more sophisticated language processing

No. If you implement these systems, the model outputs the sentence it guessed and the confidence the user said that sentence. The activation logic is tuned so it triggers on pretty low confidence so the more sophisticated system can try working on the actual command.

No, it is because the problem is much easier. Voice recognition doesn't just mean amplifying the sounds (there's only so much of that you can reasonably do), but rather discriminating a phrase from other phrases. You only have two classes: "hey siri" and everything else. This is much easier than differentiating a conversation, where words and phrases are not so easily distinguishable: "Sam went and got Tom" vs "Man went and got bomb".

To put this in an easier context, let's say we have colored shapes of various sizes. We place them down and start walking away. It's probably easier to discern color than shape at a distance (is that a hexagon or octagon?). And sufficiently far away it is hard to distinguish colors that are near one another (is that brown, firebrick, or maroon?). You can probably see that the complexity of the task is critical and that discrimination is easier when you have fewer bins to place things in.

There is a separate chip for just this, but yes.

It should be noted that this kind of spying is possible only if both the phone manufacturer and the carrier are willing (or forced) to cooperate. Otherwise, short of using social engineering to make someone install malware on their cellphone, it would be much harder to achieve. One more reason (do we really need more?) to push for Open Source in phones too, and the usual caveat emptor wrt carriers.

I’ve always found it interesting people put tape over their camera but leave the microphone untouched. If I’m concerned about privacy my words carry a lot more personal information than pictures of me picking my nose while programming.

Yeah I don't get why everyone didn't foresee this as obvious. If you're talking about something that even has a remote chance of being used against you, you put your damned phones in another room.

Putting tape over the mic wouldn't stop it from hearing you. People leave their mics untouched because no amount of touching will solve the problem without permanently damaging your device or making it useless.

Some devices will disable the internal microphone if it detects an external mic has been inserted and there are people who do things like take a microphone or a cheap pair of earbuds+mic and cut the cord off just past the jack so that they can leave it inserted without worry that it will collect anything. I've heard that some devices were updated to stop disabling the internal mic after an external one was inserted in response to this tactic, but haven't verified that myself. Either way, that's something that can sometimes be done or undone in software, so I wouldn't count on it.

I do try to avoid bugging my own house as best I can. I opened up my PS5 controller to rip out the two internal microphones for example, but there's nothing I can do to my cell phone's mic to keep it from listening while the phone is being used/useful. Removing the battery or burying the phone 10 feet below the surface of the earth or keeping it in a lead lined box makes the phone about as useful as not having one.

Tells a lot about our visual bias as human beings. Most people are able to see and identify the camera but nobody really sees the tiny hole for the microphone. Only people who know it's there think about it.

It's also funny because it's about control. Most people are exposing (probably too much of) their lives voluntarily but don't want to be exposed picking their noses.

We are only trying to be rational beings when something important depends on it or because in certain domains we've been trained to

Some laptops have physical switches for that, but phones are just kind of small and there's nowhere to put them.

Of course, if someone really cares they can put custom hardware in your phone (or anything else you carry on you.) It's all about your threat model.

I have a feeling it isn't them playing with their noses they're worried about.

With how accurate you can get remote microphones to work, much of the effort to avoid listening from government agents feels misguided. If you want to keep them from having remote capabilities, the trick isn't to try and control your surroundings, but to control them. Legislation and rules of operation are pretty much required.

You can be worried about your phone companies and such. Though, I think there is also a social problem, not a technical one.

Give me a giant hardware toggle switch please!

Again, though, if your concern is covert listening, the technology is stacked against you. Hard.

I never even use the microphone in my phone. I only talk using headsets and/or car Bluetooth. I wish I knew how to physically and permanently disable it.

Fill it with epoxy?

what's the trade-in value after that?

I'm not concerned with that. I bought the phone for $150 on ebay and intend to use it until it becomes unusable; it's a Pixel 4A 5G, the last one with a headphone jack.

Shouldn't be that hard if you have the tools to open it?

A motion sensor in your phone is also a microphone and might be able to pick up speech.

>A motion sensor in your phone is also a microphone and might be able to pick up speech.

It's also possible to read lips and monitor typing or writing motions using 802.11, so that's another modality. In addition, your phone's baseband could potentially be used for those same purposes given it affords access to an additional spectrum of RF.

I'm not sure these comments help much. The sophistication of these attacks increases a lot as well as the error/noise. Security is never infinite, it is always probabilistic. Mostly it is about putting enough roadblocks in the way of your adversary so that they don't have the resources to attack you (time, money, and/or skill). This is why the discussion of a threat scenario is always critical because that's what bounds this. I'm sure they'd have an easier time hacking your phone to get a view of your screen than it would to pull off either of the above hacks. So, removing the microphones is going to significantly hinder any adversary and lets be honest, unless you're a high profile target that's being attacked by a nation state, this will be far more than sufficient to ward off any would be attackers (including even many nation state attacks).

But as far as interesting and what is possible, these comments are that.

>I'm not sure these comments help much.

They weren't intended to. That said, I otherwise agree with everything you said.

>But as far as interesting and what is possible, these comments are that.

Yeah, it's just fun to discuss the theoretical. I'm pretty boring, so I don't personally use any exotic tinfoil precautions. Heck, if I were important I'd probably just not use a phone. Much simpler.

No worries, I mistook the intent of the comments. It is often hard to distinguish people who are stating something because it is fun and they are interested from someone who is stating something where the intent is to indicate that efforts are wasted.

> It's also possible to read lips and monitor typing or writing motions using 802.11, so that's another modality.

Doesn't that require the router to cooperate? As in, this is something that was developed to be used on purpose, but it's not very reliable on a random network you don't own.

You can see a lot of timing info from network sniffing even if you can't read it of course.

>You can see a lot of timing info from network sniffing even if you can't read it of course.

That was my primary assumption. Worst case, modern phones can act as hotspots—so if your phone is owned but stuffed in a drawer, it could theoretically sense beyond that and perhaps read your lips from another room. I'm guessing any stereotypical white vans worth their salt would provide fake devices for your phone to connect to in that case.

On the baseband side, perhaps Stingray setups have the potential to be more useful than commonly imagined.

In a few years all household appliances will be connected, including of course the already connected TVs, digital radios, IoT devices, etc. Adding a cheap and small MEMS microphone to one unused analog input of the appliance CPU would cost pennies and guarantee the manufacturers eternal "gratitude" from 3 government agencies. Think about having just to choose which device in which room to activate.

I would love to disable all those sensors

Reminds me of Snowden doing this. https://youtube.com/watch?v=ucRWyGKBVzo

Your best bet would probably be starting with an iFixit guide[0]. Even if they don't have the exact thing you want to do, some of the guides will just help you learn a lot. The first time I took part a phone I was terrified because they are expensive and I didn't have a lot of money, but I quickly learned how easy it was. It looks much harder than it is, just be patient and go slow (especially when removing the screen). Here's an example guide in how to replace the earpiece on the pixel 7 pro[1]. I chose this because it is a more involved process and because one of the microphones are in that speaker (sorry, gotta disconnect both). It may look complicated, but the number of steps is because of the thoroughness of the guide, not because of complexity. To modify this guide to your goal, all we'd need to do is find where the (other) microphones are. Which for this, that isn't hard since they have to face the outside and Google even tells us[2] (note that there are 3). Not a schematic, but we at least know the location and should be able to reason it out. What also helps in this is a teardown guide, which we can find a video version for this phone here[3], where he does mention the location of the microphones too.

I hope this helps. Just be aware that sometimes components are coupled or you may need to do some minor soldering (which in this case, is probably necessary). The good news though is that if you intend to NEVER use the microphones again, then this is much easier because if you do a little damage then it doesn't matter. To remove those surface mounted mics shouldn't be that hard, but again, patience is key. Use the largest (preferably flat tip, like a flathead screwdriver) soldering iron tip you can get in there and just push (since we're not worried about repair, you don't have to be as concerned about heat).

[0] https://www.ifixit.com/Guide

[1] https://www.ifixit.com/Guide/Google+Pixel+7+Pro+Earpiece+Spe...

[2] https://support.google.com/pixelphone/answer/7157629?hl=en#z...

[3] https://arstechnica.com/gadgets/2022/10/pixel-7-pro-teardown...

This is awesome, thank you!

I am a noob at soldering but I have done a tiny bit before. Maybe I can get this done after all!

I'd say that this is going to be one of the easiest surface mount (SMD) soldering jobs you'd ever do. Removing is easy, adding is hard. Fwiw, my first time doing SMD I was really nervous but then found it to be quite easy and so it opened up a lot of things for me (making and repairing). It is much easier than you think (often easier than through hole!), just when placing something you need to hold it in place (or use a heat gun, making these sometimes easier). Here's a short video if you want some guide[0]. As for soldering tips, any would be fine, but I learned that the flat tips, like this[1], tend to be my goto rather than the pencil like tips. The mass helps it keep heat and the larger surface area helps you transfer that heat.

[0a] Great demo, showing simplicity: https://www.youtube.com/watch?v=5uiroWBkdFY

[0b] A bit more detailed and Dave is just always extremely entertaining and a wealth of knowledge (notice his "go to" tip. Professionals use chisels ;) : https://www.youtube.com/watch?v=hoLf8gvvXXU

[1] https://www.amazon.com/UY-CHAN-Original-Soldering-Replacemen...

Hard to believe that this is considered "novel" I thought everyone knew this was fair game. It is so obvious. PinePhone and Pureism/Librem have hardware enabled switching for this reason.

"It's so obvious"

What exactly is obvious?

That every successful tech company has to play ball and give up whatever asked. If they don’t then they won’t exist for long. That everything you say and do is stored, cataloged, and searchable. Hell I’m reasonably sure nifi wouldn’t exist otherwise. There is a notable example from 2011

I always thought lavabit was the best example.

Like apple did, when they unlocked that infamous IPhone... Oh wait

Sure, the data that they collect is mostly avaible to agencies like FBA.

But the argument is about using microphone, and despite Android being fairly open platform, where you can controll every single aspect of the device all the way down to assembly, I have yet to seen the smoking gun that objectively proves that googlePlayServices, or whatever, listens to you 24/7 and transmits all that to the servers.

Apple knew they didn't need to unlock the phone. At some point a vuln was going to get discovered that would let them in, and that's exactly what happened. This way Apple gets to look like a privacy respecting shining beacon of goodness for users, while simultaneously handing over iCloud access and letting China do whatever the hell it is they do.

It was probably more novel in 2006. (I was also confused at first when they started mentioning Nokia and Razr)

> Kaplan's opinion said that the eavesdropping technique "functioned whether the phone was powered on or off." Some handsets can't be fully powered down without removing the battery; for instance, some Nokia models will wake up when turned off if an alarm is set.

It wouldn't surprise me if this is one of the big reasons manufacturers were pressured into non-removable batteries.

Not to rain on a good theory but removable batteries require extra space in the form factor and cost in manufacturing. Where smaller is a UX goal and cost is a profit goal, being forced by shady government agencies actually seems like a second order effect.

Keeping the mic live is a profit goal.

Cost is probably negligible as phones had removable batteries since the beginning, even cheap ones.

Not even a little consideration for this has made manufacturers ditch removable batteries. The primary driver has been increased capacity, followed closely by the ability to make them much thinner and develop water resistant handsets. All other reasons are so far below these they aren't even worth mentioning.

> Not even a little consideration for this has made manufacturers ditch removable batteries.

What makes you qualified to say this? Do you work for one of the large phone manufacturers?

I think some observation and a little engineering thinking makes this clear.

In the early 2010s one of the biggest issues with smartphones was battery life. The packaging of the battery is limited if a requirement is to make it removable (i.e. it must be a certain shape, have connectors in a certain location, etc). If that requirement is removed, battery can be made to fill whatever space is available in the device.

People drop their phones in water all the time. Then call the company and complain it's broken. People get mad when the company tells them it's their fault. If you have fewer holes in a device, it's less likely to be damaged by water. A removable battery means exposed connectors sealed by imperfect covers, making them more likely to be damaged by any water ingress. So they removed that failure point.

Another issue in tech generally is end-users messing with the device internals, accidentally breaking the device, and blaming the company. Remove that failure-case by making the internals inaccessible and complaints go down.

Making the device impossible to access because of a deep-state conspiracy is just a convenient mental shortcut to believe in. I'm as annoyed as the next person about less-repairable devices but I don't think it's a conspiracy, just practical engineering.

My s5 had great battery life with removable battery that could be upgraded and was perfectly waterproof almost 10 yrs ago. I think the bigger driver is encouraging people to get a new phone when battery life inevitably declined. I used the S5 for about 6 years with no issues other than a couple battery replacements. My pixel didn't even make it 2 years before serious performance declines and poor battery life.

It must have been one of the few well-made smartphones back then. It's still hard to make a perfectly waterproof seal that last for years and is inexpensive enough to use on a smartphone. Some team has to sit around designing it, so without the removable battery and seal they could in theory be freed up to do something else

Just because something is possible to do, doesn't mean every manufacturer wants to do it

I have first hand accounts from decision makers about the challenges these decisions were faced with and have worked with many engineers in adjacent areas over the years.

While I haven't worked on phones, I have worked on tablets and e-readers. I can say that the discussions have always been around cost, ID, waterproofing, tablet thickness in the design stage. I don't think it's a stretch to say the same design considerations are at work in cell phones.

I've worked with the ID and battery teams (who also worked on phones). There was never even a hint of being pushed to non-replaceable batteries from "higher up" for "reasons". Instead, it was product specs / design parameters that drove the choice.

> There was never even a hint of being pushed to non-replaceable batteries from "higher up" for "reasons". Instead, it was product specs / design parameters that drove the choice.

Cynically, we should be so lucky to have such crudely overt messaging. I would expect the reasons to take the form of specs and design parameters and not be so blunt. The NSA has done just that in the past.

https://arstechnica.com/information-technology/2013/09/the-n...

> Not even a little consideration for this has made manufacturers ditch removable batteries. The primary driver has been increased capacity, followed closely by the ability to make them much thinner and develop water resistant handsets. All other reasons are so far below these they aren't even worth mentioning.

Oh give me a break! It obviously makes more money to have things non-trivially-user-servicable - especially wear items like a battery.

I'm a strong advocate for right to repair... that being said, user serviceability is something that has to be explicitly defined as a requirement. However, just because something is difficult to service, doesn't mean it was designed with that in mind.

For example, designing a flex PCB with a 90 degree bend and a small inner radius like you might see on a mipi display or camera makes it extremely difficult to service, but has very little effect on initial assembly. If you design it for serviceability, you might do 2x 45 degree bends or a rounded pcb plus a stiffener above the mezzanine connector, reducing the chance it may tear or have broken joints.

I guess what I'm saying is just because it hasn't been designed to be serviced, doesn't mean it was designed to be unserviceable.

It generates more sales to tightly couple a wear item to its host, full stop.

We're talking about massive for-profit businesses here, there's zero chance this is happening ignorant of the positive revenue effects.

I've had to replace the battery on my phone due to swelling in the desert heat four times now.

How about the consideration that making an item less easy to service incurs more complaints? That's a metric too right?

I think you are overestimating how much influence beancounters and marketing people have on the design.

I live in the desert as well and in the last 15 years between me, my wife and my parents have only replaced a single battery. We buy a new phone once a year and then rotate them, so keep them a minimum of 3 years.

dude, you're voluntarily replacing phones too frequently for it to be relevant

try living in the desert without AC or spend more time outdoors... I tend to go through a battery per summer.

> try living in the desert without AC or spend more time outdoors

I'll pass... that sounds fucking miserable.

The EU is mandating removable batteries in consumer appliances by 2027.

Removable meaning without glue or similar adhesives, and without proprietary tools. Not removable in the when you drop your phone the lid jumps off kind of way.

So with screws then?

well, that's what I'm hoping. We'll see what manufacturer make of the regulations.

That reminds me of how the CIA was able to track Pablo Escobar. He had a mobile phone that he only turned on to make calls, however the CIA was able to remotely activate and track it using surveillance aircraft.

If conspiracy were our only worry, what a wonderful world it would be. However, dangerous developments like this happen all too often without supervillains cackling and masterminds plotting... and people make mistakes like this through innocent decisions.

We don't need to worry only about conspiracy, seemingly arbitrary design choices undermine us at every turn. No one said "if only this room-sized computer were able to fit into a speck of dust and be powered by ambient temperature differentials I could spy on everyone!"... we want those things because it is convenient. Or because it can do things we couldn't do otherwise.

The universe itself has it out for us.

Reply on news.ycombinator.com