This has been the case for quite some time (10 years or so). I've torn apart two Macbooks to specifically test this and the LED is enabled when the camera receives power. It's impossible to disable with software.

You can modify the Macbook of course, which is why the paranoid type will still use tape.

Some attacks I've seen will take a photo with the camera as fast as possible so you might not notice, but the light will always turn on.

I think newer ones even route it through the T1 chip. Need to look again.

There was an exploit, over 10 years ago like you said, that allow to disable it. But that was corrected.

Edit

- iıSeeYou: Disabling the MacBook Webcam Indicator LED - This was 2008 - https://jscholarship.library.jhu.edu/bitstream/handle/1774.2...

Looking for t1 links.

Edit 2

Can't find anything confirming the t1 right now..

The T chip in modern MacBooks is the T2, and it's certainly used for the mic cutoff. I'm unaware of public disclosures around it's use for the camera.

https://support.apple.com/guide/security/hardware-microphone...

On MacBooks the disconnect is done without the T2 being involved.

> You can modify the Macbook of course, which is why the paranoid type will still use tape.

If the attacker has done this, they might as well just hide the camera somewhere else.

Oh absolutely, but if I were targeting Zuckerberg this would now be a camera he carries with him and almost always has an available network connection (If I were a billionaire I'd be switching devices near constantly to stave off attacks). You could be carrying around the bug for at least a year if not more.

Your average blackmailer would be better off with the separate camera route (and likely wouldn't be sophisticated enough otherwise) but a state-sponsored attack could probably pull it off.

Not really - there are lots of times when you will not be present with your laptop (think airport security).

It's orders of magnitude easier to get access to a mobile device than to setup a camera in every possible location a target might move to. But they'll helpfully carry the device with them.

I really don't think airport security has enough time to open an Apple laptop and rewire the camera.

could probably do a hardware swap in 30 mins, people have definitely been detained longer than that

> It's impossible to disable with software.

> iSeeYou ( 2008 )

Which is it?

> This has been the case for quite some time (10 years or so).

even then there probably should be some minimum "flash" time in that if it turns on and off faster than the LED lights up or human eyes can see then it may still take images faster than we can register?

EDIT: a commenter below said

"Macs newer than 4 or so years also have a feature that forces the LED to stay lit for at least 3 seconds after it has been turned on to prevent the single-frame-grab attack."

Frankly, by the time the indicator is noticed, it may be too late. Even with the LED wired in place and a 'feature' (capacitor or software?) forcing it to stay lit for 3 seconds, the damage may be done. It does, at least, let you know an attack occurred.

I really think an LED is barking up the wrong tree. A built-in hardware cover would have the same effect as the LED, and no one would have to trust it, since the camera would just be recording the back of the cover if it was hacked in stealth mode.

Ironically for Americans' wariness China, the "pop-up webcam" in Huawei laptops completely resolves the security and trust issue, by providing a physical way to disable the camera.

https://www.youtube.com/watch?v=IKY3hvYKPpA

Too bad it mainly shows your nostrils :(

There are plenty of laptops with a built-in hardware cover for the camera.

The attack most people are worried about is your camera being activated while your laptop is open. Huawei's approach seems worse since you can't just apply a cover to it.

What? It's a physical button, you have to press it to open. I don't think there's a software control for it, is there?

Oh I didn't know that, I thought it popped up when the camera was activated. My mistake.

How would you handle the microphone? I don't know any hardware way to make sure my microphone's not hearing me if I can't trust the software.

A physical electrical on/off switch for the microphone (that actually interrupts the wires going to the mic).

Doesn't that require the same kind of trust in the correctness of the wiring that is at issue in some other comments? I can see that a camera cover works, but I think I couldn't see that a physical switch was actually interrupting the wiring without doing a teardown.

If you don't assume the internal wiring is as described I do not think there is a way for you to guarantee the microphone is not recording. It's even less obvious than a camera lens being covered or not, and taping over the "microphone hole" won't actually block receiving sound at all.

I do physically disconnect or cover cameras. My desktop computer has no microphone unless I plug one in. Laptop and phone though? I definitely remember in the aughts that the police could remotely turn your phone into a bug without any user indication at all. To the best of my knowledge the only way to actually be sure you aren't being recorded with your own equipment is to remove power... of course now the batteries are soldered on, totally a coincidence I'm sure.

You want your equipment to be verifiably secure?

Boy, wait until you find about what's possible manipulating PCB or processor designs (the Bloomberg fiction piece "The big Hack" illustrates what would be possible with just PCB manipulation). Or firmware or anything else proprietary running with elevated rights (normal ones are already enough to cause a lot of damage on most systems).

All you can do is reduce attack surface.

A fiction piece isn't a great demonstration of anything, really.

Well, what got it declared fiction was "a broad search doesn't turn up good evidence that this has been done", rather than "experts agree this is obviously impossible". So the reaction to the piece did demonstrate something.

> I can see that a camera cover works

Well, I don't mean to fuel your paranoia, but there are materials opaque to the human eye but not necessarily fully opaque to a CMOS sensor, especially one without infrared/UV filters.

It would be quite possible, trivial even, to design a web camera with a nice built-in black camera cover -- which is transparent to the camera.

It takes one external party checking the wiring to make sure and it can't be patched by an update.

Plug in a headphone cable that has the wire to the microphone cut. There’s also pre-made plugs on Amazon too (search “microphone blocker”).

That does nothing except switching the default audio source to the blocked channel. Any app that can access the default source can access the still working microphone just as well.

There’s apps (it’s also pretty easy to write one yourself like I did) that monitor when the input source is changed and forces it back immediately. Yes, wish that feature was built into the OS, but I also don’t grant almost any app microphone access.

Wikipedia says that doesn't work for smartphones: https://en.wikipedia.org/wiki/Microphone_blocker#Smartphone_... . If a smartphone can override it in software, then couldn't a computer also do so?

Depends on the laptop hardware in question, just as it does on which smartphone you’re using, as your link indicates.

If your laptop (or iPad, in the case of the latest gen) is an Apple T2 equipped device, this is somewhat relevant (“somewhat” as it’s not directly related to my original reply)

https://support.apple.com/en-gb/guide/security/secbbd20b00b/...

If the microphone is built into the laptop it should have a hardware switch. That still has problems as you note, but it is far better than the current state of affairs for most devices. And it still requires trust in someone (or opening the device yourself).

A better solution would be to have the microphone be physically removable, but even then it could have an internal power source and storage...

My x1 carbon has a built in webcam blocker slider built in, and it disables the hardware too when it’s slid closed.

Magnetic sliding cover: https://www.amazon.com/SpiShutter-Slim-RoseGold-Magnetic/dp/...

I used one of those (same asin) for a while. Cracked my screen just like TFA says. Thankfully Apple repaired it and looked the other way. I'm guessing they won't be so forgiving now with this article out there.

Ah... got it. Thank you for writing that.

my laptop has a piece of tape over it, but i can still see the LED lit when it comes on. however, as far as the flash of a quick snap, does the camera on the MacBook take an image faster than the iDevices? the phones always take at least a second after pushing the button, so maybe Apple would be better off putting the same kind of camera in the laptop to ensure it is impossible to take a photo quickly

If an iPhone is displaying the current camera contents on the screen, the software already has access to the camera, and is already "taking photos" in real-time.

So basically I have to check the light every 3 seconds? That seems .... unrealistic.

This is the only correct design, yet almost nobody does this. One of the challenges is verifying that this was done correctly.

You may be able to verify a design like this once with a teardown, but it's impractical to verify your particular device, and next to impossible to re-verify this every time you leave your laptop unattended.

My threat model doesn't involve someone disassembling my laptop and rewiring it, at that point they might as well just plant a bug.

This isn’t speculation. Opening up hardware to modify the internals is something that covert agencies do, and it’s naive to think that the purpose is equivalent to a simple bug.

As someone who posts to Hacker News you are by almost any definition a potential information source or APT vector.

I think what they mean is their threat model is more drive by random attacks rather than concerted efforts by intelligence agencies.

For the former a piece of tape is enough.

> This isn’t speculation.

He didn't say it was.

What he said was that his threat model doesn't include state agencies. Many people use their laptops in bed and they probably don't want random people using their laptop webcam to take random nudes.

And, while I hear the NSA being quoted a lot, let's be fair, how many of us have the opsec to stand up to a genuine NSA attack?

A reasonable goal in thwarting the NSA should be to stop some random bureaucrat from typing on a few keys and getting access to your stuff without a signature. As long as some bureaucrat has to put their signature on a piece of paper and get budget to get my stuff, I'm safe against random trawling.

Besides, if the NSA really wants you, they'll just fabricate the evidence. It's way cheaper.

> > This isn’t speculation. > He didn't say it was.

They didn’t need to use the word to imply it. I did however assume they’re weren’t so self-centeredly solipsistic as to think that their personal choices were the universal general case, or advisable for a typical participant on this forum. But then:

> Besides, if the NSA really wants you, they'll just fabricate the evidence. It's way cheaper.

Again, no. That’s another of those “surely this doesn’t happen” comments, and once again, it is flat wrong. Even setting aside that discrediting/compromising someone is likely to be different agency; just as with bugging them, there are still plenty more reasons besides eavesdropping or coercion, such as industrial espionage, or sabotage, and the main point is that this isn’t theoretical. It happens. Deal with it, don’t deal, up to you, but never assume you can’t be a target.

This is why I only take a disposable laptop and phone across US or Chinese borders. Both have intelligence services with the resources, ability, and inclination toward systematically documenting anyone even remotely interesting.

> As long as some bureaucrat has to put their signature on a piece of paper and get budget to get my stuff, I'm safe against random trawling

If this level of willingness to disregard the entire history of bureaucracy and covert action helps you feel safe, and if you have no direct or indirect ties to any other human being, body of data, or corporate entity that could ever be at any kind of risk, then why not I guess.

Apple's already addressed this by making their laptops nearly impossible to open up.

I had an Apple on-site support engineer open up a 16” MBP on my kitchen table a couple of weeks ago to replace a cracked screen, and he brought only a small tool roll, a headlamp, and a anti-static work surface and wrist-strap. Took him only a few minutes to strip it down, longer to rebuild and recommission for use.

I must add, I peeked over his shoulder a few times, and my comment that it looks as good inside as it does out was answered with “yes, it’s like being asked to restore a work of art”.

I was just trying to make a joke. I've actually been very happy with working on my MBP, it was easier to re-paste than a Dell XPS from the same year. The interior of Macbooks I've seen have been laid out remarkably well.

Yes, nearly impossible... unless you spend $5 at Amazon on a screwdriver in which case it’s trivial.

My work laptop (a Thinkpad) has a built-in plastic slider that physically covers the camera. If something activates the camera while that cover is closed, it won't see anything.

I wanted this, but I wanted the WQHD display more.

Lenovo didn't allow you to configure a T480 with both the high-resolution display and the plastic slider, it was either-or. The high-resolution display was their premium option that omitted the slider and added two more IR cameras for Windows Hello support.

So in addition to not getting the slider, I also had to disable the IR cameras.

I'm half tempted to buy the version of the display bezel with the slider and see if it'll fit as a replacement.

Not only that, but this feature is also present in X1, which is plenty thin.

I came here to post exactly this about my X1E. First laptop I've seen that has a builtin slider (though I'm sure there are others).

Bingo. It's a lot simpler to just allow the user to physically block the lens, which vetoes everything[1], rather than trust or verify that all that stuff is implemented correctly and reliably.

[1] Yes, yes, assuming you can block all frequencies the camera is capable of detecting.

The important thing isn't whether the indicator will be lit or not.

The point is whether you will be able to prevent the camera from taking a video/picture you don't want it to.

Unless you can physically disconnect or block the camera, there's not guarantee that a piece of software can't record.

The best Apple can offer here is that you will know the recording was done (assuming you are looking at your computer, and not looking away or in another room) .

A couple hypotheticals here:

A way to overvoltage/current and burn out the LED. Probability low, but nonzero.

More plausible situation: There are times when the camera is pointed at me, but my head is not pointed at the camera. Perhaps I left the laptop on in the corner of a room while it's playing music, or a myriad of other things.

A cover is a simple, foolproof solution to the camera issue.

In the case of the school district spying on the kids, the light would just blink and be gone.

Better but not bulletproof; LED's can fail. Manufacturers might consider leaving a shallow recess around the camera to accommodate the paranoid among us.

Even if they did claim that why should we trust that they did it correctly and without unintentional or intentional vulnerability?

Sure, but can you guarantee any particular webcam wasn't intercepted and modified before being delivered?

Can you guarantee the manufacturer followed the design specification for every batch?

Threat models matter.