< Return to Video

Testing Apple's Touch ID with Fake Fingerprints

  • 0:07 - 0:08
    Hey it's Norm from tested and
    I am here with:
  • 0:08 - 0:10
    Frank Bullito
  • 0:10 - 0:12
    We're here at Adam's shop
    for another video
  • 0:12 - 0:15
    Interesting video this time, it's
    an experiment.
  • 0:15 - 0:18
    yeah
    An honest to goodness experiment.
  • 0:18 - 0:20
    We have no idea how this is going to end.
  • 0:20 - 0:23
    So today we're going to talk about technology.
    Are you comfortable?
  • 0:23 - 0:25
    ok..
    Special effects and technology
  • 0:25 - 0:28
    Last year apple released the iPhone 5-S,
  • 0:28 - 0:30
    which was their first product
    using 'Touch I.D.'
  • 0:31 - 0:33
    You've probably heard of Touch I.D.,
    it's a bio-metric sensor system
  • 0:34 - 0:36
    for your phone, to double as
    your password.
  • 0:37 - 0:39
    Now this year there are more products
    using Touch I.D.
  • 0:39 - 0:40
    and more services using them,
  • 0:41 - 0:44
    it's being used for things like Apple-pay,
    passwords, one password
  • 0:44 - 0:48
    And we wanted to see, basically
    how fool-proof it was.
  • 0:48 - 0:53
    We're not going to hack Touch I.D
    people have made claims about hacking it
  • 0:53 - 0:57
    Last year a hacking group over
    in Germany
  • 0:57 - 0:59
    named 'Chaos Computer Club' released
    a video
  • 0:59 - 1:01
    showing how they 'spoofed'
    Touch I.D
  • 1:02 - 1:04
    and what they did was they duplicated
    someone's fingerprint
  • 1:04 - 1:08
    and put it on acetate and created a
    relief image,
  • 1:08 - 1:11
    and it was a pretty involved process.
  • 1:11 - 1:14
    But we wanted to-
    We wanted to make it more involved
  • 1:15 - 1:18
    Even more involved, but basically try
    to understand how Touch I.D works.
  • 1:18 - 1:20
    Um, do you know how this works?
  • 1:20 - 1:21
    Why don't you explain.
    (laughs)
  • 1:21 - 1:26
    So beneath this home button
    on the iPhone and the new iPads
  • 1:26 - 1:29
    is a sensor, basically an optical sensor,
    almost like a camera.
  • 1:29 - 1:33
    There is a capacitance sensor,
    which makes the touch screen work,
  • 1:33 - 1:35
    so it knows, it registers
    that your finger is there.
  • 1:35 - 1:41
    But basically it takes a high resolution image,
    people have said it's about 500 dpi.
  • 1:41 - 1:46
    Which is fairly high, Apple talks about
    sub-dermal imaging, but basically,
  • 1:46 - 1:51
    it knows all the grooves and all
    the curves of your fingerprint
  • 1:51 - 1:54
    and registers that, uh because you do
    need to register it multiple times,
  • 1:54 - 2:01
    stores it locally, so when you put your finger
    on there it recognises that it is your fingerprint.
  • 2:01 - 2:06
    So we wanted to see if we could activate,
    unlock a phone, activate Touch I.D
  • 2:06 - 2:10
    without having the actual hand there;
    a warm bodied hand,
  • 2:10 - 2:12
    maybe a.. dismembered hand.
    There you go. (laughs)
  • 2:12 - 2:17
    So this is actually a hand-cast we did
    a couple of months ago, of my hand
  • 2:18 - 2:21
    It's just a silicone hand-cast and
    based on your experience in the
  • 2:22 - 2:26
    special effects industry Frank,
    how accurate is the fingerprints
  • 2:26 - 2:28
    on something like a hand-cast?
  • 2:28 - 2:31
    Well, I mean, the silicone doesn't
    shrink or expand or anything,
  • 2:31 - 2:34
    so it's pretty accurate, as long
    as we got in all the detail
  • 2:34 - 2:37
    This was done in an alginate mold,
    so the detail might be a little soft
  • 2:37 - 2:40
    but what we'll do today is we'll do a
    little silicone cast of your finger
  • 2:40 - 2:43
    ..or my finger and we'll see if
    we can crack it with that.
  • 2:43 - 2:46
    So we're talking about mold materials
    and casting materials
  • 2:46 - 2:50
    and we're using silicone as a mold material
    and in doing it right-
  • 2:51 - 2:52
    You're going to get a more accurate casting.
  • 2:53 - 2:57
    Hopefully it will be almost exactly 1-to-1,
    y'know.. microns of accuracy.
  • 2:57 - 2:58
    Hopefully.
    Hopefully.
  • 2:58 - 3:02
    And then the casting material, also important.
    yeah
  • 3:02 - 3:05
    'cause you have resins and silicones
    and all sorts of stuff.
  • 3:05 - 3:08
    Well the pad of your fingertip is about
    a 35-A durometer, but-
  • 3:08 - 3:12
    When you say Durometer that's how
    stiff it is, how rigid it is
  • 3:12 - 3:14
    Yeah, the Durometer scale is how
    you measure how soft
  • 3:14 - 3:17
    or how rigid something is.
    There's an 'A' scale for soft things,
  • 3:17 - 3:19
    and a 'D' scale for rigid things.
  • 3:19 - 3:23
    And your thumb-tip is about a 35-A.
    ok
  • 3:23 - 3:26
    But, the problem is is that when
    we push on a button
  • 3:26 - 3:29
    it doesn't have enough give to
    kinda like.. flatten out,
  • 3:30 - 3:32
    so I think we want to go a
    little bit softer.
  • 3:32 - 3:36
    So we were doing this off-camera,
    this casting is done in
  • 3:36 - 3:41
    a 20 durometer silicone, which technically
    is softer than the 35 of your finger.
  • 3:41 - 3:43
    I think we want to go softer than that,
    I have a 10 Durometer.
  • 3:44 - 3:47
    That's right, you want it to be able
    to flatten without pushing the button.
  • 3:47 - 3:49
    Yeah.
  • 3:50 - 3:52
    Now the other piece of
    the puzzle is capacitance.
  • 3:52 - 3:54
    You need to be able to activate that
    touch sensor in the first place.
  • 3:55 - 3:59
    How are we going to give silicone capacitance?
    Because if you put a silicone finger,
  • 3:59 - 4:01
    or you know, a hot-dog on here,
    it's not going to activate the touch-screen.
  • 4:03 - 4:05
    Maybe a hot-dog will... would a-
    A hot-dog would, yes.
  • 4:05 - 4:07
    But a silicone hot-dog wouldn't.
    We can make hot-dog fingers.
  • 4:07 - 4:08
    uh.. we could do that.
  • 4:09 - 4:13
    I think, from what everybody's told me,
    adding graphite is enough to
  • 4:13 - 4:17
    give it capacitance. So that's mixing it into
    the silicone or powdering
  • 4:17 - 4:20
    the surface of the mold, and then
    putting the silicone behind it
  • 4:20 - 4:24
    so that the silicone grabs the
    surface of the graphite.
  • 4:24 - 4:25
    We'll try it.
    We'll try it.
  • 4:25 - 4:29
    Like I said, this is an honest-to-goodness
    experiment and-
  • 4:29 - 4:31
    And there's a good likelihood we'll fail.
    (laughs)
  • 4:31 - 4:35
    But follow along the process,
    let's go to step one
  • 4:35 - 4:38
    which is creating a casting of
    your finger.
  • 4:38 - 4:43
    If you're using 'body double' from 'Smooth-on,'
    which is just a casting silicone,
  • 4:43 - 4:47
    Specifically for life casting.
    For life casting, it's accurate as possible,
  • 4:47 - 4:51
    we've done life-casts with it before,
    you mix in a little bit of material as well?
  • 4:51 - 4:56
    I mix in a thickener called '5X,' instead
    of using Cabosil or something fumey
  • 4:56 - 5:00
    to thicken it, it's just a couple drops
    of this liquid and it gets thick.
  • 5:00 - 5:04
    And then you're, in terms of making the mold,
    you're just doing a thin coating
  • 5:04 - 5:08
    of silicone, just to get into those grooves
    and make sure it's a strong relief,
  • 5:08 - 5:12
    a perfect copy, and then just
    slather that silicone on top.
  • 5:12 - 5:14
    Yeah. I kinda smeared it into
    the grooves of my fingertip.
  • 5:14 - 5:16
    And then built it up so that
    it has some thickness.
  • 5:16 - 5:19
    Alright, so now it's time to
    wait until this cures.
  • 5:20 - 5:24
    Ok Frank, while the silicone mold
    on your finger is curing,
  • 5:24 - 5:27
    Let's begin experiment 2.
  • 5:27 - 5:30
    Another way we thought we could
    'spoof' Touch I.D. is
  • 5:30 - 5:33
    if someone left their fingerprint on
    a piece of glass for example,
  • 5:33 - 5:40
    and we'd captured it photographically.
    So we had you put you finger right here
  • 5:40 - 5:44
    on some baby powder and put it
    right on this piece of foam-core.
  • 5:44 - 5:49
    This is basically the equivalent of putting
    your greasy finger on glass, for example.
  • 5:49 - 5:55
    Which is what the 'C.C.C.' did.
    So, we took a photo of this,
  • 5:55 - 5:58
    took a macro photo, blew it up,
    very high resolution
  • 5:58 - 6:02
    with my camera and then
    processed it on a computer
  • 6:02 - 6:05
    to scale it, to be exactly the size
    of your finger.
  • 6:05 - 6:09
    Measured it, it was about an inch tall.
    And then I laser etched it.
  • 6:09 - 6:14
    So I tried several variations, laser
    etched of different relief.
  • 6:14 - 6:16
    With laser etching y'know-
    Different depths.
  • 6:16 - 6:19
    Yeah. Different depths, changed the colour,
    inverted it, uhh but basically
  • 6:19 - 6:26
    used this as a substitute for your finger
    and then mold on top of that,
  • 6:26 - 6:30
    or cast on top of that. We're not sure
    if that's going to work either,
  • 6:30 - 6:32
    but we also want to create
    the molds here.
  • 6:32 - 6:34
    And also using silicone.
    Sure.
  • 6:34 - 6:37
    I ran some 'Body Double' onto
    these etched parts,
  • 6:37 - 6:41
    to see how they come out and some of them
    look like they're actually going to work.
  • 6:41 - 6:43
    So what we want to do is
    we want to dust them with graphite.
  • 6:44 - 6:46
    And then, put some more silicone
    on top of this, and then
  • 6:47 - 6:48
    we can use it as kind-of
    a finger pad,
  • 6:48 - 6:53
    You know, because it's thin and flexible
    and maybe that.. that will work.
  • 6:53 - 6:58
    So we're going to do some mental
    gymnastics here,
  • 6:58 - 7:03
    the photo of your fingerprint, which
    I blew up and high-contrasted,
  • 7:03 - 7:06
    made sure that all the curves
    were very visible.
  • 7:07 - 7:10
    That was the reverse of
    what your fingerprint actually is.
  • 7:10 - 7:15
    So without inverting it, without flipping
    it back, just etching that,
  • 7:16 - 7:19
    basically, what you're going to get
    on the etch is what you would get
  • 7:19 - 7:23
    on the foam-core here or a
    fingerprint on the phone.
  • 7:23 - 7:28
    And so when we create the silicone
    casting on top of that,
  • 7:28 - 7:30
    this isn't the mold, this is the fingerprint.
    Yes.
  • 7:30 - 7:35
    Ok. So put some graphite on there
    and we'll mix up some more 'Body Double'
  • 7:35 - 7:39
    And we'll add a little graphite
    into the 'Body Double'
  • 7:39 - 7:42
    Just to help with capacitance maybe, I dunno.
  • 7:51 - 7:53
    It's black like tar!
  • 7:55 - 7:58
    So the cast of your finger is about done.
    Yeah, this is all solid now so..
  • 7:59 - 8:02
    What you do is just kind of
    peel it off.
  • 8:02 - 8:04
    Because it's in all the detail of my
    finger it's got a lot of suction.
  • 8:05 - 8:06
    That's good.
    Yeah.
  • 8:07 - 8:12
    Take a look at that, you can flip it inside-out
    and see all those ridges and grooves,
  • 8:13 - 8:15
    There you go Internet, copy this
    and go hack all my passwords.
  • 8:16 - 8:18
    That's a good looking fingerprint.
  • 8:18 - 8:22
    So what we'll do is put a few
    different materials into this.
  • 8:22 - 8:27
    But, we have another copy of the fingerprint
    etched at a different DPI
  • 8:28 - 8:30
    and we'll smear some more stuff into that.
    Yeah.
  • 8:30 - 8:35
    My question and my concern with the
    laser etching is the resolution you can etch.
  • 8:36 - 8:42
    And you can change that on the laser cutter,
    so I have 500ppi, 600ppi and 800ppi.
  • 8:42 - 8:46
    ppi or dpi?
    It's.. sort of the same, points per inch.
  • 8:46 - 8:47
    ok.
  • 8:47 - 8:54
    And the higher we go the more chance
    that the acrylic will get fused into the crevices
  • 8:54 - 8:59
    and the dust will take away some of the detail,
    but that- you can recognise it as a fingerprint
  • 8:59 - 9:03
    I just don't know if it will be
    good enough to fool Touch I.D.
  • 9:03 - 9:06
    You don't know if we're good enough to fool Touch I.D?
  • 9:06 - 9:08
    Are we smarter than Apple?
    Probably not.
  • 9:20 - 9:24
    Alright, back to experiment 1,
    the hand-cast - the finger cast
  • 9:24 - 9:27
    and you're going to put some silicone in there.
    Yeah.
  • 9:27 - 9:29
    So first thing I'm going to do is
    hit it with a little bit of release
  • 9:29 - 9:32
    This is just 'Ease-release 200.'
  • 9:32 - 9:34
    We actually haven't seen you use
    mold remover before,
  • 9:34 - 9:36
    this extends the life of your molds.
  • 9:37 - 9:39
    Yeah, if you want to cast like Urethane
    parts or something like that
  • 9:40 - 9:43
    this works great, it just happens
    to be what I have sitting around right now.
  • 9:44 - 9:48
    I want to put release on here because
    nothing sticks to silicone but other silicone,
  • 9:48 - 9:50
    but we're putting silicone in here so..
    Oh..
  • 9:51 - 9:54
    So I'm spraying it in and just
    sort of rubbing it around
  • 9:57 - 10:04
    So what I'm going to do on this first
    is I'll rub some graphite into the surface
  • 10:04 - 10:06
    kind of like how I dusted the etching.
  • 10:06 - 10:09
    You've gotta be careful, this is really tricky
    because we want just enough graphite
  • 10:10 - 10:16
    to activate Touch I.D, and the activation
    I think happens around the ring of the home button
  • 10:16 - 10:21
    But we don't want too much graphite, because
    that might take away the detail.
  • 10:21 - 10:25
    I want to Imagine Apple engineers watching
    this and just laughing their butts off.
  • 10:26 - 10:30
    'Those fools, those morons,
    they'll never do it.'
  • 10:33 - 10:36
    So this is the rubber I'm going to use,
    the 10 Durometer,
  • 10:36 - 10:38
    for this stuff we were using the 'Body Double'
    because it was real thin and
  • 10:38 - 10:41
    it'll just serve as a pad you
    put on your finger.
  • 10:41 - 10:46
    So this one- I'll make kind of a full finger,
    sort of like this stuff.
  • 10:46 - 10:51
    And I'll do another one that's just
    kind of a pad we can put on another finger.
  • 11:05 - 11:08
    Alright, so the castings are done.
    The silicone has cured,
  • 11:08 - 11:11
    It's time to pop 'em out of the molds.
    Yep.
  • 11:11 - 11:14
    So what we have here,
    we actually did more than 1.
  • 11:14 - 11:20
    A silicone casting from a silicone mold,
    pops out and how do the fingerprints look?
  • 11:20 - 11:26
    I think that the detail is really great, it's
    about as exact of a replica of a fingerprint
  • 11:26 - 11:29
    as you can get from your fingerprint, but
    we'll see if it reads on the Touch I.D.
  • 11:30 - 11:33
    And then also the silicone we used on top
    of the laser etching,
  • 11:33 - 11:38
    we pulled those out as well. And the
    advantage of these is that
  • 11:38 - 11:42
    if someone wanted to 'spoof' someone elses
    fingerprint, this is more likely
  • 11:42 - 11:45
    how they would be able to do it. They're not
    going to be able to get someone to
  • 11:46 - 11:48
    put their finger in a mold for
    15-20 minutes.
  • 11:48 - 11:50
    They're more likely to take a photo
    of their fingerprint,
  • 11:50 - 11:52
    from a touchpad or piece of glass
    or something,
  • 11:52 - 11:55
    and see if they can laser etch it.
    Yeah.
  • 11:55 - 11:58
    Plus these are thinner.
    Yeah, that'll help with the capacitance.
  • 11:58 - 12:03
    It's like if you're wearing a rubber glove
    you can still activate your phone.
  • 12:03 - 12:07
    So let's do a first test which is
    a test of capacitance.
  • 12:07 - 12:12
    Let's unlock this phone,
    I'll just type in the passcode..
  • 12:12 - 12:20
    And show that using the graphite infused
    silicone works, just to manipulate the touch screen.
  • 12:21 - 12:23
    Alright, that seems to be working.
  • 12:24 - 12:32
    Let's try with this now, this is Franks finger.
    Oh, no capacitance.
  • 12:33 - 12:38
    Umm.. we could try wetting it or..-
    Putting some graphite on top of it.
  • 12:38 - 12:40
    Yeah we could try those two things.
  • 12:42 - 12:45
    This is tough because water
    just beads off silicone.
  • 12:48 - 12:52
    Ok, a wet finger,
    a wet finger-casting.
  • 12:52 - 12:54
    That activates the screen.
    Mm-hmm.
  • 12:54 - 12:58
    But we don't know how that
    will affect Touch I.D.
  • 12:58 - 13:02
    So this one's dusted with extra graphite.
    We'll see if that works.
  • 13:03 - 13:05
    It does!
    Great.
  • 13:05 - 13:09
    So capacitance is working, so here
    comes the moment of truth.
  • 13:09 - 13:11
    Or failure.
    Or failure.
  • 13:11 - 13:12
    (laughs)
  • 13:12 - 13:15
    The moment of failure,
    potential failure. We'll call it that.
  • 13:16 - 13:19
    I'll hit the top button on the screen,
    it's going to want to unlock.
  • 13:20 - 13:24
    And we'll show that, for example,
    my finger doesn't unlock this.
  • 13:26 - 13:31
    Your finger does unlock.
    My finger that was molded does unlock.
  • 13:34 - 13:39
    And now let's put that on top of there.
    (gasp)
  • 13:39 - 13:41

    It unlocked!
    It unlocked!
  • 13:42 - 13:46
    Holy moly!
    We fuckin' did it!
  • 13:46 - 13:48
    I'm genuinely shocked, I didn't
    think it was going to work.
  • 13:49 - 13:51
    Let's see if we can duplicate that.
    Alright..
  • 13:54 - 13:59
    Let's do- let's prove that this
    finger here does not work.
  • 14:01 - 14:05
    Try again
    Does not work, ok.
  • 14:05 - 14:08
    (gasp!)
    We did it.
  • 14:09 - 14:12
    (laughs)
    Huge success.
  • 14:13 - 14:15
    Test 1, experiment 1;
    Success.
  • 14:16 - 14:18
    I'm going to try it on my finger.
  • 14:23 - 14:25
    So let's talk about what did work.
    Ok.
  • 14:25 - 14:31
    Experiment 1: we did a silicone casting of your
    finger, with a silicone mold
  • 14:31 - 14:34
    that you carefully made, also of
    the same index finger.
  • 14:34 - 14:38
    Now with the silicone casting we infused
    it with graphite, which by itself
  • 14:38 - 14:43
    didn't seem to activate the capacitance ring
    on the Touch I.D sensor.
  • 14:44 - 14:49
    But by brushing graphite on top of it,
    that actually made it register as a finger.
  • 14:49 - 14:55
    And, with one of our castings we were
    able to unlock the phone.
  • 14:56 - 15:01
    And the cooler thing even, is that on new
    iPhones 6 and 6+, is theoretically
  • 15:02 - 15:06
    more improved, the sensor on the
    inside is higher resolution.
  • 15:06 - 15:12
    And even on the 6+, where we trained beforehand,
    your index finger but not mine...
  • 15:13 - 15:14
    It unlocks!
  • 15:15 - 15:17
    So now I can unlock
    Will's phone any time I want.
  • 15:18 - 15:21
    Well.. if your fingerprint was stored in it, yes.
    (laughs)
  • 15:21 - 15:24
    But the other thing we have to test of course
    is the laser etching process.
  • 15:25 - 15:31
    This is taking a high resolution photograph
    left on a surface, laser etching it
  • 15:31 - 15:33
    and then creating a
    mold of that.
  • 15:34 - 15:36
    So we've cut out a bunch of our
    laser etched fingers,
  • 15:37 - 15:41
    Let's go back to the 5-S, ok,
    and let's see if that works.
  • 15:43 - 15:48
    So here I have your fingerprint
    theoretically, on top of mine..
  • 15:51 - 15:54
    ..'try again.'
  • 15:56 - 15:58
    Come on!..... do something!..
  • 16:00 - 16:07
    So unfortunately, the laser etched and then
    cast silicone fingerprint doesn't work.
  • 16:08 - 16:14
    What's interesting is that it does recognise
    on the Touch I.D, it is trying a fingerprint.
  • 16:14 - 16:17
    Which, if you put any other capacitive surface on there-
    Like a hot-dog.
  • 16:18 - 16:21
    Right. It knows that that's not a finger,
    so it wouldn't even try.
  • 16:22 - 16:25
    So we got it to try, we got it to fail,
    but we couldn't get it to
  • 16:25 - 16:28
    recognise it as your finger.
    Well..
  • 16:28 - 16:31
    At least we got one of them.
    At least we got one of them.
  • 16:32 - 16:36
    This is really cool, we get to learn basically
    how Touch I.D. works in practice.
  • 16:36 - 16:40
    We recognise it doesn't need a warm finger.
    Mmhmm.
  • 16:40 - 16:45
    All you need is something that has
    the fidelity, the resolution of your fingertip
  • 16:45 - 16:49
    and has enough capacitance to activate the ring.
    Yeah.
  • 16:49 - 16:53
    So if your finger was chopped off,
    theoretically, it would still work.
  • 16:54 - 16:55
    There you go.
    There you go.
  • 16:56 - 16:59
    Touch I.D. tested, it's a fun project
    with Frank Bellino
  • 17:00 - 17:02
    What else do you do? You do
    a bunch of other tests here or-
  • 17:02 - 17:05
    Well, during the week, when I was at lunch
    and stuff like that, I tried a bunch
  • 17:05 - 17:09
    of different things. One thing I did was
    take a ball of clay and just push
  • 17:09 - 17:12
    my finger into it, to get a fingerprint in the clay,
    and put silicone into there.
  • 17:13 - 17:17
    That didn't work either, these are attempts at
    just doing a fingerprint in clay.
  • 17:17 - 17:22
    That's not enough resolution, the silicone
    mold gets a better resolution.
  • 17:22 - 17:24
    So that's kind of the way to do that.
  • 17:24 - 17:28
    Try different ways and you know what?
    If you have experience with special effects
  • 17:28 - 17:34
    or casting.. or new ideas.. the Choas
    Compter Club, they posted their instructions
  • 17:34 - 17:40
    we'd love to see your videos or your attempts
    at 'spoofing' Touch I.D.
  • 17:40 - 17:42
    But this was a lot of fun!
    Yeah it was cool
  • 17:42 - 17:47
    I've never- I've always thought about this but
    never really tried it, but now we kind of know
  • 17:47 - 17:49
    sort of what works.
    It's a fun experiment.
  • 17:49 - 17:52
    Touch I.D tested here on Tested.
    Thank you Franky Belito,
  • 17:53 - 17:55
    Thank you to premium members for
    bringing this video
  • 17:55 - 17:57
    And thanks 'Smooth on' for giving
    us stuff to fool with.
  • 17:58 - 18:02
    Thanks 'Smooth on' for materials again, we used
    'Body Double,' we used 'Dragon skin,'
  • 18:02 - 18:07
    and they have a lot of materials. It's fun stuff
    and you'll find more of this stuff on Tested,
  • 18:07 - 18:11
    Subscribe to our Youtube channel
    and we'll see you next time. bye.
Title:
Testing Apple's Touch ID with Fake Fingerprints
Description:

more » « less
Video Language:
English
Duration:
18:18

English subtitles

Revisions