WEBVTT - How deepfakes are impacting elections

0:00:00.520 --> 0:00:04.400
<v Speaker 1>Already and this is the Daily This is the Daily OS.

0:00:05.120 --> 0:00:06.840
<v Speaker 1>Oh now it makes sense.

0:00:14.640 --> 0:00:17.360
<v Speaker 2>Good morning and welcome to the Daily OS. It's Sunday,

0:00:17.400 --> 0:00:19.640
<v Speaker 2>the second of March. I'm Billy, I'm Zara.

0:00:19.920 --> 0:00:23.560
<v Speaker 1>We're back in your ears on a Sunday morning again. Again.

0:00:23.720 --> 0:00:27.120
<v Speaker 2>We love surprising you with a new episode on Sundays.

0:00:27.160 --> 0:00:29.080
<v Speaker 1>And Zara, I think today's topic.

0:00:28.960 --> 0:00:33.800
<v Speaker 2>Is very relevant and interesting and super interesting. So we

0:00:33.880 --> 0:00:37.120
<v Speaker 2>are nearly, at any moment now about to enter election

0:00:37.240 --> 0:00:38.600
<v Speaker 2>season in Australia.

0:00:38.760 --> 0:00:41.280
<v Speaker 3>The number of times that we have said that ONCAS

0:00:41.520 --> 0:00:45.760
<v Speaker 3>we're nearly there, We're just there. However, many synonyms can

0:00:45.840 --> 0:00:48.040
<v Speaker 3>say for we are close to an election.

0:00:48.320 --> 0:00:49.600
<v Speaker 1>We are close to an election.

0:00:49.880 --> 0:00:52.120
<v Speaker 2>So there is one that is due to be held

0:00:52.240 --> 0:00:55.400
<v Speaker 2>on the seventeenth of May, or before if Prime Minister

0:00:55.480 --> 0:00:58.720
<v Speaker 2>Anthony Alberanezi decides for us to go to an election earlier,

0:00:59.080 --> 0:01:01.560
<v Speaker 2>and that could mean that we are about to enter

0:01:01.640 --> 0:01:05.880
<v Speaker 2>a period of dis and misinformation in Australia. So today

0:01:06.040 --> 0:01:09.160
<v Speaker 2>we are talking about a very specific type of false

0:01:09.200 --> 0:01:10.720
<v Speaker 2>information and that.

0:01:10.760 --> 0:01:11.759
<v Speaker 1>Is deep fakes.

0:01:12.040 --> 0:01:14.920
<v Speaker 2>More specifically, we want to talk about how deep fakes

0:01:15.000 --> 0:01:18.720
<v Speaker 2>can impact elections, and in this context we're talking about

0:01:18.760 --> 0:01:22.520
<v Speaker 2>deep fakes that are deceptive and harmful. A quick note

0:01:22.600 --> 0:01:26.360
<v Speaker 2>before we start. This podcast is produced with financial support

0:01:26.440 --> 0:01:31.280
<v Speaker 2>from Microsoft. However, the content does remain entirely independent. And

0:01:31.360 --> 0:01:34.560
<v Speaker 2>if you're wondering why Microsoft would be interested in sponsoring

0:01:34.640 --> 0:01:36.360
<v Speaker 2>a podcast on deep fakes.

0:01:36.080 --> 0:01:37.319
<v Speaker 4>In it, I was quite interested.

0:01:37.480 --> 0:01:40.000
<v Speaker 2>Well you, yeah, because you might not know that they

0:01:40.040 --> 0:01:43.360
<v Speaker 2>are very involved in this space and they have teams

0:01:43.360 --> 0:01:46.560
<v Speaker 2>that track the threat of actors who are often the

0:01:46.600 --> 0:01:48.680
<v Speaker 2>ones creating deep fakes.

0:01:49.040 --> 0:01:52.160
<v Speaker 3>Very interesting, now, Billy, you have used the word deep

0:01:52.160 --> 0:01:55.240
<v Speaker 3>fake a record number of times in the introduction of

0:01:55.320 --> 0:01:58.600
<v Speaker 3>a podcast. For anyone who's not familiar with the term, though,

0:01:58.600 --> 0:02:00.960
<v Speaker 3>what are you talking about when you say deep fake?

0:02:01.320 --> 0:02:05.160
<v Speaker 2>So they are digital photos or videos of someone that

0:02:05.400 --> 0:02:09.880
<v Speaker 2>looks very real, but they have actually been created using

0:02:10.000 --> 0:02:14.880
<v Speaker 2>artificial intelligence and they are fake, although often they're not

0:02:15.240 --> 0:02:18.240
<v Speaker 2>entirely fake. So it could be a video where ninety

0:02:18.280 --> 0:02:21.440
<v Speaker 2>percent of that video is real, but then there's ten

0:02:21.520 --> 0:02:25.640
<v Speaker 2>percent that is generated by AI and is fake. But

0:02:25.680 --> 0:02:28.280
<v Speaker 2>it's that ten percent that is really deceptive.

0:02:28.520 --> 0:02:29.280
<v Speaker 4>Well, I guess yeah.

0:02:29.320 --> 0:02:31.680
<v Speaker 3>When you're looking at something and so much of it

0:02:31.720 --> 0:02:34.120
<v Speaker 3>looks real, your sense of what's real and what's fake

0:02:34.200 --> 0:02:34.480
<v Speaker 3>is so.

0:02:34.480 --> 0:02:37.880
<v Speaker 2>Distortive exactly, and that is a key feature of deep fakes,

0:02:37.880 --> 0:02:42.639
<v Speaker 2>that they are highly highly realistic. Now, one famous example

0:02:42.639 --> 0:02:44.880
<v Speaker 2>that comes to my mind when we talk about deep

0:02:44.880 --> 0:02:46.880
<v Speaker 2>fakes is I don't know if you've seen this there

0:02:47.040 --> 0:02:49.359
<v Speaker 2>or if you remember this, but the image of Pope

0:02:49.360 --> 0:02:53.600
<v Speaker 2>Francis in a Valenciago quilted jacket. It's kind of walking

0:02:53.720 --> 0:02:58.120
<v Speaker 2>the streets in this massive white jacket, and it went

0:02:58.240 --> 0:03:01.480
<v Speaker 2>completely viral at the time in March train twenty three,

0:03:01.680 --> 0:03:04.200
<v Speaker 2>and a lot of people obviously didn't know it was

0:03:04.240 --> 0:03:07.480
<v Speaker 2>a fake photo because it looked so real, and two

0:03:07.600 --> 0:03:11.280
<v Speaker 2>years ago, there wasn't really the conversations happening at the

0:03:11.280 --> 0:03:13.560
<v Speaker 2>time that there are now, when we are much more

0:03:13.600 --> 0:03:16.920
<v Speaker 2>aware that there are so many fake images, and obviously

0:03:17.240 --> 0:03:19.799
<v Speaker 2>we are talking about AI so much even we're talking

0:03:19.800 --> 0:03:21.680
<v Speaker 2>about AI so much more today than we were a

0:03:21.680 --> 0:03:24.600
<v Speaker 2>month ago, let alone two years ago, and so that

0:03:24.760 --> 0:03:26.640
<v Speaker 2>is one of I would say that's one of the

0:03:26.639 --> 0:03:31.000
<v Speaker 2>more innocuous examples of deep fakes where although it was fake,

0:03:31.080 --> 0:03:34.600
<v Speaker 2>it didn't necessarily do any damage by people believing that

0:03:34.720 --> 0:03:37.320
<v Speaker 2>it was real. But why we're talking about it today

0:03:37.440 --> 0:03:41.040
<v Speaker 2>is because deep fakes can and they have been used

0:03:41.040 --> 0:03:44.520
<v Speaker 2>in a much more deceptive way, and one of those

0:03:44.560 --> 0:03:46.600
<v Speaker 2>ways is during elections.

0:03:46.920 --> 0:03:48.400
<v Speaker 4>I see how you're bringing this altogether.

0:03:49.360 --> 0:03:50.240
<v Speaker 1>We're circling that.

0:03:50.960 --> 0:03:54.960
<v Speaker 2>So there have been examples around the world where politicians

0:03:55.080 --> 0:03:58.200
<v Speaker 2>or political candidates or even just people who want a

0:03:58.200 --> 0:04:02.000
<v Speaker 2>certain party to win, have used deep fakes to mislead

0:04:02.080 --> 0:04:05.400
<v Speaker 2>voters and to benefit their campaign. Or it could not

0:04:05.520 --> 0:04:07.560
<v Speaker 2>even be for that reason, it could just be to

0:04:07.920 --> 0:04:09.760
<v Speaker 2>cause chaos or cause disruption.

0:04:10.360 --> 0:04:12.840
<v Speaker 3>Okay, So we're talking about it because we're in the

0:04:12.880 --> 0:04:16.160
<v Speaker 3>lead up to an election and we've seen not just

0:04:16.400 --> 0:04:19.560
<v Speaker 3>here but across the world examples of deep fakes being

0:04:19.720 --> 0:04:23.080
<v Speaker 3>used during an election cycle. Can you talk me through

0:04:23.080 --> 0:04:26.520
<v Speaker 3>some of the examples that we've seen of this and

0:04:26.560 --> 0:04:27.599
<v Speaker 3>how it's been used.

0:04:27.839 --> 0:04:28.000
<v Speaker 4>Yeah.

0:04:28.000 --> 0:04:31.520
<v Speaker 2>I thought we would start in Australia because you might

0:04:31.560 --> 0:04:34.320
<v Speaker 2>not be aware that there have actually been examples in

0:04:34.440 --> 0:04:38.560
<v Speaker 2>Australian elections or by Australian parties of using deep fakes.

0:04:38.920 --> 0:04:42.839
<v Speaker 2>So Queensland, Hm'm sure you've heard of it, familiar. They

0:04:43.000 --> 0:04:46.800
<v Speaker 2>had an election last year and the Liberal National Party

0:04:47.000 --> 0:04:51.400
<v Speaker 2>posted a video of the then Premier, Steven Miles dancing

0:04:51.800 --> 0:04:55.000
<v Speaker 2>with text beneath it that said pov my rent is

0:04:55.080 --> 0:04:57.400
<v Speaker 2>up sixty dollars a week, my power bill is up

0:04:57.440 --> 0:05:00.720
<v Speaker 2>twenty percent. But the premier made a sandwich on TikTok,

0:05:00.960 --> 0:05:04.000
<v Speaker 2>So basically they were saying that the premiere at the

0:05:04.000 --> 0:05:06.720
<v Speaker 2>time was dancing and was happy despite the fact that

0:05:07.000 --> 0:05:09.799
<v Speaker 2>the rent in the state had gone up on average

0:05:09.800 --> 0:05:13.760
<v Speaker 2>for the average renter. Now, that video was actually labeled

0:05:13.920 --> 0:05:17.240
<v Speaker 2>as AI generated, although I would say that the label

0:05:17.440 --> 0:05:20.880
<v Speaker 2>was pretty easy to miss if you weren't necessarily looking

0:05:20.880 --> 0:05:22.880
<v Speaker 2>out for it, and so it was just this completely

0:05:22.920 --> 0:05:25.200
<v Speaker 2>fake video of Steven Male dancing.

0:05:26.000 --> 0:05:28.839
<v Speaker 3>I remember seeing that at the time and coming up

0:05:28.880 --> 0:05:30.880
<v Speaker 3>on you for you page it did, and I remember

0:05:30.920 --> 0:05:35.159
<v Speaker 3>thinking that it was the first time I recall seeing

0:05:35.200 --> 0:05:36.880
<v Speaker 3>something like that here in Australia.

0:05:37.080 --> 0:05:40.280
<v Speaker 2>Yeah, and it generated a lot of media coverage. But

0:05:40.360 --> 0:05:42.920
<v Speaker 2>then a lot of that coverage then pointed out that

0:05:43.320 --> 0:05:46.560
<v Speaker 2>this is not unique to one side of politics. And

0:05:46.800 --> 0:05:50.600
<v Speaker 2>actually around that time, well actually before that video was posted,

0:05:50.920 --> 0:05:54.240
<v Speaker 2>the Australian Labor Party had also posted a video on

0:05:54.279 --> 0:05:57.760
<v Speaker 2>their TikTok that had used AI to create a fake

0:05:57.839 --> 0:06:01.440
<v Speaker 2>video of Federal Opposition leader Peter dune and dancing with

0:06:01.520 --> 0:06:04.440
<v Speaker 2>the words dance if you want to build nuclear power

0:06:04.480 --> 0:06:07.600
<v Speaker 2>plants in everyone's backyard. Zara, I'm going to send you

0:06:07.680 --> 0:06:08.920
<v Speaker 2>the video for you to look.

0:06:09.000 --> 0:06:09.520
<v Speaker 1>Right now.

0:06:10.320 --> 0:06:12.680
<v Speaker 2>I realized this is an audio platform and so the

0:06:12.760 --> 0:06:15.200
<v Speaker 2>audience can't see it. But Zara, I want to get

0:06:15.200 --> 0:06:15.760
<v Speaker 2>your reaction.

0:06:16.120 --> 0:06:19.280
<v Speaker 3>My god, oh my god.

0:06:19.520 --> 0:06:23.120
<v Speaker 2>I'm going to say, it's probably not the most realistic.

0:06:23.240 --> 0:06:25.200
<v Speaker 4>It looks nothing like Peter Dudden.

0:06:25.000 --> 0:06:27.640
<v Speaker 2>Okay, but you can see that it's trying to kind

0:06:27.680 --> 0:06:31.680
<v Speaker 2>of imitate him vaguely. Sure, Look, this is one of

0:06:31.720 --> 0:06:35.400
<v Speaker 2>the examples where it is pretty clear to most people.

0:06:35.200 --> 0:06:36.400
<v Speaker 1>That it was a fake video.

0:06:36.760 --> 0:06:39.640
<v Speaker 3>However, it must be said that like you and I

0:06:39.760 --> 0:06:42.440
<v Speaker 3>look at Peter Dudden day and day out where we

0:06:42.560 --> 0:06:44.960
<v Speaker 3>work in news, we're very familiar with what he looks like.

0:06:45.000 --> 0:06:48.200
<v Speaker 3>I imagine for someone that's perhaps not paying the same amount

0:06:48.240 --> 0:06:50.159
<v Speaker 3>of attention, maybe they wouldn't be able to tell.

0:06:50.040 --> 0:06:53.160
<v Speaker 2>You yes exactly. And also, I think the Stephen Miles one,

0:06:53.200 --> 0:06:55.039
<v Speaker 2>if you look at that one, it was a lot

0:06:55.120 --> 0:06:59.520
<v Speaker 2>more realistic. Yeah, but again, these are videos of the politicians,

0:07:00.000 --> 0:07:03.600
<v Speaker 2>and so it is one of the more benign examples

0:07:03.640 --> 0:07:06.640
<v Speaker 2>of how deep fakes can be used in elections.

0:07:07.240 --> 0:07:10.080
<v Speaker 3>Okay, So if you're saying that those are benign examples,

0:07:10.160 --> 0:07:12.960
<v Speaker 3>or perhaps more benign examples. What are some of the

0:07:12.960 --> 0:07:16.120
<v Speaker 3>more troublesome deceptive examples that we've got out there.

0:07:16.280 --> 0:07:19.560
<v Speaker 2>Yeah, so if we look overseas, there are quite a

0:07:19.600 --> 0:07:24.640
<v Speaker 2>few more troublesome examples. So one example that the audience

0:07:24.720 --> 0:07:26.440
<v Speaker 2>might be familiar with or might have seen in their

0:07:26.440 --> 0:07:29.720
<v Speaker 2>feeds is that during the US election last year, in

0:07:29.760 --> 0:07:33.040
<v Speaker 2>the lead up to that election, Elon Musk shared a

0:07:33.160 --> 0:07:36.960
<v Speaker 2>video that looked like an ad campaign for Kamala Harris,

0:07:37.240 --> 0:07:41.000
<v Speaker 2>who was the Democratic candidate for the US presidency last year.

0:07:41.360 --> 0:07:42.560
<v Speaker 2>Here's a bit of that clip.

0:07:43.040 --> 0:07:46.680
<v Speaker 5>I Kamal Harris and your Democrat candidate for presidents because

0:07:46.800 --> 0:07:50.400
<v Speaker 5>Joe Biden finally exposed to sinilitive to be thanks, John,

0:07:50.760 --> 0:07:53.520
<v Speaker 5>I was selected because I am the ultimate diversity hire.

0:07:53.600 --> 0:07:56.080
<v Speaker 5>I'm both a woman and a president of color. So

0:07:56.160 --> 0:07:59.600
<v Speaker 5>if you've criticized anything I say, you're both sexist and racist.

0:08:00.120 --> 0:08:03.880
<v Speaker 3>That one's crazy because that sounds exactly like her.

0:08:04.160 --> 0:08:07.119
<v Speaker 1>Yeah, it would be look video.

0:08:07.400 --> 0:08:10.160
<v Speaker 2>Yeah, if you were watching it, it would be really

0:08:10.200 --> 0:08:13.120
<v Speaker 2>hard to tell that that is a fake video. And

0:08:13.200 --> 0:08:16.960
<v Speaker 2>you probably couldn't immediately tell unless you were in tune

0:08:17.000 --> 0:08:19.880
<v Speaker 2>with every single thing Kamla Harris has ever said, you

0:08:19.880 --> 0:08:22.080
<v Speaker 2>would have to do some digging to figure out if

0:08:22.080 --> 0:08:25.160
<v Speaker 2>that was real or fake, especially when it's shared by

0:08:25.400 --> 0:08:30.280
<v Speaker 2>a very influential person. Now, that video received nearly one

0:08:30.360 --> 0:08:35.080
<v Speaker 2>hundred and forty million views according to x's Metrics, and

0:08:35.360 --> 0:08:38.840
<v Speaker 2>it was liked by nearly one million accounts. Wow, so

0:08:38.880 --> 0:08:41.200
<v Speaker 2>that gives you an idea of how viral this kind

0:08:41.280 --> 0:08:42.520
<v Speaker 2>of content can go.

0:08:43.080 --> 0:08:46.120
<v Speaker 4>I'm just looking at the tweet responses. What's that called

0:08:46.240 --> 0:08:47.360
<v Speaker 4>yesh x response?

0:08:47.400 --> 0:08:51.079
<v Speaker 3>And ex responds someone who was retweeted one hundred and

0:08:51.120 --> 0:08:54.080
<v Speaker 3>ninety three times and like ten thousand times said.

0:08:54.000 --> 0:08:55.160
<v Speaker 4>Is this AI or reel?

0:08:55.400 --> 0:08:55.640
<v Speaker 3>Yeah?

0:08:55.880 --> 0:08:57.280
<v Speaker 4>So clear genuinely no idea.

0:08:57.440 --> 0:09:00.920
<v Speaker 2>Yeah, it clearly wasn't that clear to me any many people.

0:09:01.120 --> 0:09:03.120
<v Speaker 2>As I'm sure you've guessed. The reason we're talking about

0:09:03.160 --> 0:09:06.720
<v Speaker 2>it is because it was completely fake. Kamala Harris never

0:09:06.800 --> 0:09:11.000
<v Speaker 2>said those words, so that voice was completely AI generated,

0:09:11.120 --> 0:09:16.360
<v Speaker 2>but it sounded and looked extremely realistic. Now, it's hard

0:09:16.440 --> 0:09:19.959
<v Speaker 2>to measure what impact a video like that had on

0:09:20.040 --> 0:09:22.520
<v Speaker 2>the minds of voters in the US. It could have

0:09:22.640 --> 0:09:27.199
<v Speaker 2>had no impact at all, or it could be quite significant, Again,

0:09:27.360 --> 0:09:30.880
<v Speaker 2>especially when shared by a very influential person such as

0:09:31.120 --> 0:09:35.000
<v Speaker 2>Musk So that's one example, but there are also examples

0:09:35.080 --> 0:09:40.079
<v Speaker 2>of foreign countries using deep fakes to interfere with elections.

0:09:40.280 --> 0:09:41.600
<v Speaker 4>Can you unpack that more for me?

0:09:41.800 --> 0:09:44.760
<v Speaker 2>Yeah, it's quite a big concept to get your head around.

0:09:44.800 --> 0:09:50.120
<v Speaker 3>So foreign actors interfering with domestic elections in a different country.

0:09:49.960 --> 0:09:52.560
<v Speaker 2>Exactly, And it's not really something that many of us

0:09:52.559 --> 0:09:55.120
<v Speaker 2>would ever think about, but it's probably something that you've

0:09:55.240 --> 0:09:58.560
<v Speaker 2>heard murmurs about, this idea that there are countries who

0:09:58.600 --> 0:10:02.720
<v Speaker 2>try to interfere with other countries' elections with the aim

0:10:02.880 --> 0:10:06.559
<v Speaker 2>of electing a government they believe will be more closely

0:10:06.640 --> 0:10:09.840
<v Speaker 2>aligned with their own policies and agendas, or like I

0:10:09.880 --> 0:10:12.120
<v Speaker 2>said before, it could not be because of that, it

0:10:12.160 --> 0:10:16.920
<v Speaker 2>could be again just to create chaos and disruption to society.

0:10:17.240 --> 0:10:18.400
<v Speaker 4>What are some examples of that?

0:10:18.800 --> 0:10:20.840
<v Speaker 2>So if we stick with looking at the US election

0:10:20.960 --> 0:10:24.920
<v Speaker 2>last year, Microsoft actually did a report into how rushing

0:10:24.960 --> 0:10:29.800
<v Speaker 2>groups created deep fakes to try to influence the US election. Now,

0:10:29.840 --> 0:10:31.760
<v Speaker 2>before I go on, I do just want to quickly

0:10:31.800 --> 0:10:35.360
<v Speaker 2>mention that, as we discussed earlier, this podcast is sponsored

0:10:35.400 --> 0:10:38.280
<v Speaker 2>by Microsoft, but when not required to talk about them

0:10:38.320 --> 0:10:41.080
<v Speaker 2>in this section, they just have done a lot of

0:10:41.120 --> 0:10:43.920
<v Speaker 2>research in this area that I thought was very relevant

0:10:43.920 --> 0:10:47.240
<v Speaker 2>to talk about. Okay, good clear disclaimer, just want to

0:10:47.240 --> 0:10:50.520
<v Speaker 2>be completely transparent there. So in a report last year,

0:10:50.760 --> 0:10:54.000
<v Speaker 2>Microsoft found that two groups that were aligned with the

0:10:54.080 --> 0:10:59.000
<v Speaker 2>Russian government, the Kremlin, had created and disseminated videos that

0:10:59.120 --> 0:11:03.040
<v Speaker 2>were quote designed to discredit Harris again talking about Kamala

0:11:03.080 --> 0:11:07.920
<v Speaker 2>Harris and stoke controversy around her campaign. So, for example,

0:11:07.960 --> 0:11:11.599
<v Speaker 2>they found this one video that showed Harris supporters attacking

0:11:11.720 --> 0:11:14.880
<v Speaker 2>an attendee at a Trump rally, and that video was

0:11:15.000 --> 0:11:19.000
<v Speaker 2>completely false, but it was still seen by millions and

0:11:19.120 --> 0:11:22.760
<v Speaker 2>millions of people, and there was no labeling or anything

0:11:22.920 --> 0:11:24.400
<v Speaker 2>that indicated.

0:11:23.800 --> 0:11:25.000
<v Speaker 1>This video was fake.

0:11:25.440 --> 0:11:28.680
<v Speaker 2>So to someone who was just on their feedscrolling through,

0:11:29.040 --> 0:11:32.840
<v Speaker 2>there wasn't any clear signs that this was a fake video.

0:11:33.520 --> 0:11:35.880
<v Speaker 2>And so basically what they found is that these groups

0:11:36.000 --> 0:11:40.160
<v Speaker 2>intentionally created these videos and planted them in front of

0:11:40.320 --> 0:11:45.720
<v Speaker 2>US audiences to sway their opinion in this case against Harris.

0:11:46.120 --> 0:11:48.600
<v Speaker 3>Yeah, it's really interesting and again needs to be said,

0:11:48.720 --> 0:11:51.040
<v Speaker 3>this isn't happening on one side of politics. We've seen

0:11:51.080 --> 0:11:55.319
<v Speaker 3>examples across the aisle when it comes to this Billy

0:11:55.360 --> 0:11:57.240
<v Speaker 3>one thing that I think about when you're talking about

0:11:57.240 --> 0:12:01.319
<v Speaker 3>this is the onus that exists on on the individual,

0:12:01.520 --> 0:12:05.520
<v Speaker 3>because ultimately we're saying you can be served up fake

0:12:05.640 --> 0:12:09.720
<v Speaker 3>content that is mimicking reality but is not real, and

0:12:09.800 --> 0:12:13.360
<v Speaker 3>sometimes deceptively, sometimes in a more benign nature, and we

0:12:13.400 --> 0:12:17.280
<v Speaker 3>are expecting audiences to be able to critically, I guess,

0:12:17.320 --> 0:12:22.040
<v Speaker 3>analyze what they're seeing and understand how can people tell

0:12:22.120 --> 0:12:24.480
<v Speaker 3>if videos that they're watching or photos they're seeing a

0:12:24.600 --> 0:12:27.240
<v Speaker 3>real or AI generator. You know, for example, we've seen

0:12:27.280 --> 0:12:30.280
<v Speaker 3>the President of the United States share something recently there

0:12:30.360 --> 0:12:32.920
<v Speaker 3>was completely AI generated and completely fake.

0:12:33.160 --> 0:12:33.360
<v Speaker 4>Yeah.

0:12:33.360 --> 0:12:35.199
<v Speaker 2>I speak to my friends about this all the time

0:12:35.280 --> 0:12:39.120
<v Speaker 2>because we're always, you know, sending each other videos and whatnot,

0:12:39.240 --> 0:12:41.480
<v Speaker 2>and we're always talking about how do we tell if

0:12:41.480 --> 0:12:44.439
<v Speaker 2>this is real or not? And what's hard is that

0:12:44.640 --> 0:12:48.920
<v Speaker 2>by definition, deep fakes are incredibly realistic. The whole point

0:12:48.960 --> 0:12:49.760
<v Speaker 2>is for you to think.

0:12:49.640 --> 0:12:50.440
<v Speaker 1>That they are real.

0:12:51.200 --> 0:12:53.800
<v Speaker 2>I think the main thing to look for is who

0:12:53.840 --> 0:12:56.480
<v Speaker 2>is sharing this video and what is the primary source

0:12:56.600 --> 0:12:59.199
<v Speaker 2>or does a primary source even exist? So if it's

0:12:59.240 --> 0:13:03.720
<v Speaker 2>from an verified, completely random account, then that's more likely

0:13:03.800 --> 0:13:07.040
<v Speaker 2>to be a fake video than one shared by a

0:13:07.080 --> 0:13:12.240
<v Speaker 2>reputable news outlet, or even better, a primary source, so Zara,

0:13:12.360 --> 0:13:15.120
<v Speaker 2>as you know, at TDA, we are all about the

0:13:15.160 --> 0:13:18.360
<v Speaker 2>primary source, So that means looking at where each and

0:13:18.440 --> 0:13:21.679
<v Speaker 2>every piece of information originates from. So just for some

0:13:21.720 --> 0:13:24.840
<v Speaker 2>context for the listeners, our guiding rule at TDA is

0:13:24.840 --> 0:13:28.200
<v Speaker 2>that every fact we share needs to have a primary source.

0:13:28.480 --> 0:13:31.439
<v Speaker 2>We need to know exactly where it came from and

0:13:31.880 --> 0:13:34.960
<v Speaker 2>go to that place to get that information ourselves.

0:13:35.360 --> 0:13:36.800
<v Speaker 1>So if we look at the video of.

0:13:36.880 --> 0:13:39.560
<v Speaker 2>Kamala Harris shared by Elon Musk, that kind of is

0:13:39.559 --> 0:13:43.520
<v Speaker 2>more murky because obviously Elon Musk's page is verified, so

0:13:43.840 --> 0:13:45.880
<v Speaker 2>I understand why a lot of people would have thought

0:13:45.880 --> 0:13:48.560
<v Speaker 2>it was real. But again, if we are looking for

0:13:48.640 --> 0:13:51.280
<v Speaker 2>the primary source in that case, you would find it

0:13:51.320 --> 0:13:55.600
<v Speaker 2>on either Harris's own verified page, all the Democratic parties

0:13:55.720 --> 0:13:59.480
<v Speaker 2>verified socials. And if you can't find that video on either,

0:14:00.000 --> 0:14:02.800
<v Speaker 2>and there's a good chance it's not real. But I

0:14:02.800 --> 0:14:06.720
<v Speaker 2>think you can also consider whether that content perfectly fits

0:14:06.720 --> 0:14:10.640
<v Speaker 2>a particular narrative being pushed by someone that it potentially

0:14:10.720 --> 0:14:13.880
<v Speaker 2>could be too good to be true. And I also

0:14:13.920 --> 0:14:16.280
<v Speaker 2>just want to say that this isn't about making you

0:14:16.360 --> 0:14:19.480
<v Speaker 2>think that everything is fake. That is not what we're saying.

0:14:19.720 --> 0:14:22.840
<v Speaker 2>I think we're just talking about developing a healthy level

0:14:22.880 --> 0:14:27.400
<v Speaker 2>of skepticism, not too much that you completely distrust everything

0:14:27.440 --> 0:14:30.480
<v Speaker 2>you see, but enough to stop and consider where this

0:14:30.600 --> 0:14:34.120
<v Speaker 2>content is from. Could it be fake or manipulated? And

0:14:34.200 --> 0:14:37.000
<v Speaker 2>if you're basing your opinion on this video that you're seeing,

0:14:37.520 --> 0:14:40.240
<v Speaker 2>should you be checking whether there is a primary source?

0:14:40.760 --> 0:14:42.640
<v Speaker 3>Now, Billy, I just want to end back. I guess

0:14:42.640 --> 0:14:45.720
<v Speaker 3>where we started with a discussion about it here in Australia.

0:14:45.800 --> 0:14:48.400
<v Speaker 3>Are there laws that regulate this sort of behavior here?

0:14:48.760 --> 0:14:51.440
<v Speaker 2>Yeah, our audience has asked us a lot about this,

0:14:51.800 --> 0:14:54.640
<v Speaker 2>and essentially the answer is that there is nothing that

0:14:54.720 --> 0:14:59.120
<v Speaker 2>prohibits the use of AI in election campaigns. Also remembering

0:14:59.160 --> 0:15:03.200
<v Speaker 2>that this is quite new technology. There is, however, regulation

0:15:03.360 --> 0:15:07.640
<v Speaker 2>about campaign communication, so you know, information coming from the

0:15:07.720 --> 0:15:11.280
<v Speaker 2>Labor Party directly or the Liberal Party or whatever party

0:15:11.360 --> 0:15:15.600
<v Speaker 2>or political candidate, that communication is regulated. But as we

0:15:15.640 --> 0:15:18.600
<v Speaker 2>saw in many of the examples above, the deep fakes

0:15:18.600 --> 0:15:23.200
<v Speaker 2>that are misleading voters aren't usually coming from political parties directly.

0:15:23.240 --> 0:15:27.240
<v Speaker 2>They're coming from other third party actors who perhaps have

0:15:27.400 --> 0:15:31.320
<v Speaker 2>a certain agenda to push. It is a criminal offense

0:15:31.360 --> 0:15:35.160
<v Speaker 2>in the Electoral Act to mislead or deceive an elector.

0:15:34.800 --> 0:15:38.000
<v Speaker 1>When it comes to casting a vote. But again, this.

0:15:37.960 --> 0:15:41.160
<v Speaker 2>Becomes harder to apply to videos posted on random accounts

0:15:41.160 --> 0:15:44.040
<v Speaker 2>where the exact source of that video or image is

0:15:44.080 --> 0:15:44.520
<v Speaker 2>not known.

0:15:45.040 --> 0:15:49.000
<v Speaker 3>Look a really complex area, and especially as it interacts

0:15:49.040 --> 0:15:51.520
<v Speaker 3>with election season. Definitely one that we have to keep

0:15:51.560 --> 0:15:53.920
<v Speaker 3>an eye on and one that will continue to keep

0:15:53.960 --> 0:15:54.600
<v Speaker 3>speaking about.

0:15:54.640 --> 0:15:55.840
<v Speaker 4>So thank you for taking us through that.

0:15:55.840 --> 0:15:56.560
<v Speaker 1>Billy, Thank you.

0:15:56.800 --> 0:15:59.120
<v Speaker 3>We'll be back again tomorrow morning with a deep dive

0:15:59.240 --> 0:16:02.000
<v Speaker 3>as usual, but until then, enjoy your weekend.

0:16:06.120 --> 0:16:08.480
<v Speaker 1>My name is Lily Maddon and I'm a proud Arunda

0:16:08.680 --> 0:16:11.240
<v Speaker 1>Bungelung Kalkadin woman from Gadighl Country.

0:16:12.040 --> 0:16:15.200
<v Speaker 2>The Daily oz acknowledges that this podcast is recorded on

0:16:15.240 --> 0:16:17.720
<v Speaker 2>the lands of the Gadighl people and pays respect to

0:16:17.800 --> 0:16:21.120
<v Speaker 2>all Aboriginal and torrest Rate island and nations. We pay

0:16:21.160 --> 0:16:24.080
<v Speaker 2>our respects to the first peoples of these countries, both

0:16:24.120 --> 0:16:25.040
<v Speaker 2>past and present.