YT YouTube2Text

YouTube2Text → Transcripts → Technology | SABC News anchors used in deep fake videos: Moshoeshoe Monare

Technology | SABC News anchors used in deep fake videos: Moshoeshoe Monare — Transcript

by SABC News · 2,960 words · 413 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00and now as mentioned earlier let's go to
0:01this conversation there have been
0:02several videos during the rounds on
0:04social media purporting to be of some
0:06sabc news anchors advertising a project
0:09by Elon Musk now these ads are running
0:12on Facebook and YouTube mostly now these
0:14deep fakes claim you can make as much as
0:1630,000 Rand a day and sometimes even
0:18more and that some prominent business
0:20people have already invested in this but
0:23if you haven't seen them have a
0:26look breaking news Elon musk's project
0:29has has alarmed the government and major
0:31Banks he has introduced a new secret
0:34investment that has made hundreds of
0:36people very rich last week he appeared
0:39in public and announced his new project
0:41which according to him can turn anyone
0:44into a millionaire within five to six
0:46months without any difficulties and with
0:48minimal investment of
0:50$250 the Innovative businessman urged
0:53everyone to seize this incredible
0:56opportunity and of course a few minutes
0:58after the interview ended the government
1:00called and requested that the interview
1:02not be broadcast but it was too late I
1:05am providing access to my platform for
1:07my fellow countrymen where I come from
1:10to get the system up and running a
1:12minimum investment of $250 is required
1:15this is the minimum amount at which the
1:17system will start making profitable
1:19trades in the stock and cryptocurrency
1:21markets to gain access you need to
1:23register on the official website wait
1:25for a manager's call who will assist you
1:27in the subsequent steps including
1:29account funding and program setup no
1:31longer need to work such a statement was
1:33made by Elon Musk as he opens access to
1:36his new development which generates
1:38money automatically using artificial
1:40intelligence as a Trader uh I am
1:42providing access to my platform for my
1:44fellow countrymen where I come from to
1:46get the system up and running a minimum
1:48investment of 4,700 Rand is required
1:51this is the minimum amount at which the
1:53system will start making profitable
1:54trades in the stock and cryptocurrency
1:56markets to gain access you need to
1:59register on the official website wait
2:01for a manager's call who will assist you
2:03in the subsequent steps including
2:05account funding and program
2:08setup and we can tell you that these
2:11videos are fake uh we certainly are not
2:14as anchors even endorsing this we don't
2:16even know what it's about but let's have
2:17this conversation with sabc group
2:19executive for news and current affairs M
2:21who is joining us in studio thank you so
2:23much for your time your reaction to the
2:25videos cuz you've been seeing a lot of
2:27them coming from us look my my reaction
2:29is that um it's scary but uh of the most
2:34concerning aspect of it is the fact that
2:37they are using my colleagues faces and
2:40voices knowing
2:42that because they are from an
2:45institution or a broadcaster with some
2:48credibility where there is public trust
2:51and my worri is one is that um anything
2:53that then that public trust is affecting
2:56us asbc news is affecting the anchors
2:59who who are seen as leading journalist
3:01who are seen as um you know um whatever
3:05they express that there is should be
3:07some legitimacy to it and the most
3:10worrying thing I think is the fact that
3:11we are in the era where we we we we we
3:15want to see what is the good and um you
3:18know the most benefiting about
3:19artificial intelligence but the worest
3:22one is that if artificial intelligence
3:24is going to be abused this way um then
3:27it really makes people to be skeptical
3:29especially and the fact that there will
3:30be the good and the positives that we
3:32need to um Embrace and and and and and
3:36and and accept and look as as we see um
3:40news is that it's worrying because you
3:42know for the first time when it happens
3:44with one of our anchors on morning life
3:46it's something that we investigated and
3:48the most um difficult thing is the
3:52inability to trace the inability to
3:55trace the roots and the source and I
3:57think that's the most dangerous thing
3:58because they keep on
4:00popping up and they are using or abusing
4:04the vulnerability of members of the
4:07public who are desperate to invest and
4:09seeing the most credible public figures
4:12like yourself um Elon Musk and other
4:15colleagues and they think this is this
4:17should be true and my worries one is
4:19that also it Dent our credibility
4:21because our brand is something that as a
4:24public broadcaster something that people
4:26trust and I wonder then um you know
4:29because it's so difficult to trace what
4:32do we do in in in such a situation as a
4:35as a public
4:36broadcaster look like I said it's a
4:38difficult one because if you were able
4:39to trace we will um actually block it I
4:42remember um times ago um even before AI
4:47um there used to be this kind of like um
4:50scams coming through programmatic
4:52advertising on um our previous online
4:55Publications and you would block that
4:58particular um
5:00um you know advertising and it will just
5:04pop up in a different manner look the
5:06only thing that we we can do is to go
5:08public about it tojust ens show and um
5:12say to members of the public you know as
5:14a public broadcaster or employees of
5:16public broadcaster it's against our
5:19editorial policy it's against our public
5:21mandate to be advertising products
5:25especially this kind of products on
5:26behalf of people that actually we you
5:29know it's it's just to educate it's just
5:31to go on a on campaigns that really tell
5:34um our audience to say we do not get
5:36involved in this kind of things and
5:38finally then let's reiterate the message
5:41to someone at home who's sitting and is
5:43probably thinking or has not even seen
5:46it what do they do do they invest do
5:48they not what does the public
5:49broadcaster say directly to them the
5:52sabc news will say to them we do not
5:55propagate or we are not agencies or
5:58agents of any product or any institution
6:02that really says You must invest in this
6:05product or that product you know we are
6:07in the business of news and the um what
6:10we do is based on credibility on
6:13accuracy on um seeking the truth and on
6:17the trust that we have built um
6:19throughout the years between us and the
6:21public and we would want to restore that
6:23trust to say some of the videos that you
6:25see where our employees or anyone
6:28associated with us saying that invest in
6:31that is not our policy and it's against
6:33our editorial policy to call for people
6:36to invest in in in such um schemes so
6:40you've heard it first here again um
6:43we're saying do not invest do not invest
6:46in that the same way as Leanne of course
6:49has been at pains to describe what has
6:51been done to her as well do not invest
6:53Iman is now the latest one who has been
6:57used as well to try and sell this
6:58particular scam purporting to be by Elon
7:01Musk so you've heard it from us do not
7:04invest please make sure that your
7:06hardend money is not lost to these
7:08criminals but let me thank you so much
7:09for your time M that is sabc group
7:11executive for news and current affairs M
7:14and man is standing by to talk to an AI
7:17expert about all of this and how it gets
7:20done thanks buir so let's get that
7:22specialist opinion for you now and
7:24joining us is Lavina Ramis who is also
7:27an advisor on all things AI uh related
7:31for the African Union amongst others
7:33Lina thank you very much for your time
7:35this is really upsetting it's uh scary
7:37and as you heard from our group head of
7:39news there the questions being asked how
7:41do we Trace its roots and
7:44Source uh that's such a pertinent
7:46question and uh firstly apologies and
7:49you know my heart goes out to anyone uh
7:52you know including uh the likes of the
7:54sabc families that have been affected
7:57you know in terms of these kind of
7:59scenarios so you know to answer the
8:01question I guess it's really around how
8:03do we actually verify information and
8:05how do we verify misinformation or how
8:08do we verify the single source of truth
8:10right um and this is such a prival role
8:13that I think a media can play so we've
8:15seen some interesting plays uh across
8:18different media hazards um from simple
8:20things like a verification badge or a
8:23simple Watermark uh sort of scenario
8:26that really depicts to is it uh content
8:29that is created by humans versus is it
8:31actually created by
8:33AI it's very hard I think you know for
8:36for those of us outside of the domain of
8:39AI the languaging and uh you know the
8:41technology Associated to be able to know
8:44what to look out for we talk about
8:46watermarks and verification badges um
8:49just share with our viewers in plain
8:51terms where and how they can find this
8:54out yeah absolutely so I think um you
8:57know let's start with just simple things
9:00of a a tone so you know really look out
9:03for the voice in a message and really
9:06pay attention to is it as authentic uh
9:09or is it not as authentic if you get an
9:11initial feel uh that something's not
9:14right uh simply just don't engage um
9:17that's the first so voice the second is
9:19from an image perspective really look to
9:21tell if the image is race still like so
9:25you know if that is very much the case
9:28then uh that's generally a good
9:29indication that it is AI generated um
9:33you know other sort of aspects that you
9:35can look out for is make sure that you
9:38verify the links right uh just don't
9:41click on something um you know if you
9:44receive an email make sure that the
9:46email address that you receive it from
9:48is actually legitimate itself um you
9:52know you've got to take your due
9:53necessary process um you know from a
9:56user perspective and uh take take back
9:58that control and make sure as well um
10:02whatever information you're actually
10:03sharing on your web or on any sort of U
10:07you know digital platform you actually
10:09are aware what you're sharing at any
10:11given point in time let make it really
10:13simple Lina I'm so sorry to jump in
10:15there because when you talk about
10:17verifying the links what are the you
10:20notable things you should look out for
10:22in that email some of them really mimic
10:25official credible emails quite
10:27well yeah absolutely so things like uh a
10:31proper email address we generally know
10:34you know companies have uh either a. Za
10:37at the end of it with their uh sort of
10:40company name in the middle uh so that
10:42should be like a standard uh you know
10:45also if you're not sure of the domain or
10:47their email uh domain Google it right um
10:50go and have a look if uh their email
10:53address says info@ sabc do. Za as an
10:57example uh you know then you compare it
11:00to that particular email address that
11:02you're then seeing as as one example the
11:05second would be um you know people that
11:08request for any personal information
11:10just simply never engage never divulge
11:14whether it is from a bank whether it is
11:15from an insurance company a Telco um
11:18literally anyone right that's the first
11:21the second I agree with everyone here
11:23that really stipulated just simply don't
11:25invest if anyone asks you for any type
11:28of funding simply don't invest um and I
11:31think those are simple things that you
11:33could really do uh and you can never be
11:35over overly cautious yeah there two
11:38things I really want to get to hopefully
11:39there's time that I think are important
11:41that don't only focus on how this
11:44technology facial recognition technology
11:46can be used um to scam in this way um
11:50the one is what is this technology that
11:52makes us so possible in stuff that we
11:54read we think about Clear View for
11:56example cited in a couple of Articles
11:58around um their capabilities for AI
12:01scraping in other words whatever your um
12:04internet or online presence is they can
12:06create a composite by scraping
12:08everything that's related to your ID and
12:10build up a profile on you which also can
12:13you know can potentially play into into
12:15these scam so very quick word on this
12:18technology this type of
12:20Technology yeah it is something called
12:22Deep fake technology and uh that
12:24literally uses a bit of um you know
12:27computer visioning which is part of AI
12:30as well as speech recognition and NLP uh
12:34and in a nutshell you know um it is
12:37something that uh can be very
12:39detrimental we've seen lots of cases uh
12:41that actually depict to that um yeah so
12:44from a technology perspective I would
12:46actually also say the reverse is true as
12:48much as it's detrimental why are we not
12:50actually seeing enough companies or AI
12:53companies coming up that actually helped
12:55to prevent uh you know these sort of
12:57cyber attacks or deep fake tax 24/7 in
13:00terms of web scraping absolutely 100% uh
13:04that is something that I I think policy
13:06makers are working really hard at
13:08actually you know uh getting rid off
13:10completely and uh you know as you said
13:13it actually does more detriment uh than
13:15anything else so you know a user and
13:17your thread of information becomes
13:19really important if you've automatically
13:22signed on using your Gmail account or
13:24any other sort of account make sure that
13:26you go in and un uh you log out as an
13:30example um you know just never leave
13:32your Wi-Fi on all the time connected to
13:35your PC as an example there's just
13:38simple little things that you've got to
13:39just uh really take back full control so
13:42much more that we've got to be
13:43hypervigilant around and to close
13:45reading and the daily Maverick did a
13:47really great article in this uh this
13:48could also have potentially devastating
13:51consequences uh you know for something
13:53as simple as being in a job interview
13:55the recruiter cross references your
13:56image from a Google search sees
13:59potentially what your political and
14:00other affiliations are and potentially
14:02not give you the job if it conflicts
14:04with their own beliefs and this stuff
14:06talks about how invaded Our Lives can
14:10become for this you know because of this
14:11technology on the other hand yes it can
14:14help law enforcement it can track and
14:15Trace people around the world from a
14:17traffic light to a store they might be
14:19purchasing Goods in but when it comes to
14:22this Invasion on our our private lives
14:24and and you know us appearing uh in
14:27places in other people's photographs
14:28even um that becomes really quite
14:31dangerous and
14:33invasive yeah I think um you know that
14:36intimate balance that we need to find
14:38between man and machine uh is really
14:40really important um there's something
14:43called uh your you know your digital uh
14:45space uh pretty much similar to your
14:48personal space uh in the physical world
14:50and uh that is becoming more and more uh
14:53you know uh important and we've seen you
14:56know the likes of uh a lot of what
14:59happens in the physical world replicated
15:01in the digital world and that's you know
15:03yeah so we kind of really need to
15:06balance that uh to our best ability and
15:09I think this is why the trend thus far
15:11is really headed towards empowering the
15:13user and giving the user consent to
15:16understand in real time uh you know only
15:19for now for this Split Second I want to
15:21share my bank details with my bank so
15:23that I can get pre-approved loan as an
15:26example post that pre-approved
15:29once that's done the bank doesn't know
15:31that I exist right so that's pretty much
15:34where we need to be getting to and and
15:36driving TOS in 30 seconds how can we
15:38take this down our colleagues uh you
15:40know all of our faces are emblazoned
15:42across these um these little you know
15:44these scams what can we do to get rid of
15:47them yeah that's a that's a tough
15:50question um I think it is obviously
15:52contacting the websites directly uh you
15:55know requesting for those sort of things
15:56to happen uh forming uh you know some
15:59sort of uh a Consolidated voice that
16:02then uh you know goes to the likes of
16:04The Regulators to then assist in
16:07actually you know bringing that down um
16:10yeah unfortunately that in itself is not
16:12a very easy task I mean um I could be
16:16really you know um you know play with
16:18things at this point in time and say
16:20maybe create an alternative AI but I
16:22won't because that would be too
16:24controversial um but yeah I mean uh I
16:27think for now it's just you know try to
16:30be as human as possible uh try to
16:32connect as much uh you know try to
16:34reason as much and um yeah I'm I'm sure
16:37somewhere along the line we'll be able
16:39to uh take all of this down so much
16:42again left uh in this conversation that
16:43we'd love to explore perhaps another
16:45time thank you so much for yours
16:47technologist for Africa and Technology
16:49advisor to the African Union Lavina
16:57ramun

About this transcript

This page contains the full transcript of Technology | SABC News anchors used in deep fake videos: Moshoeshoe Monare by SABC News, generated from the public captions YouTube serves with the video. The transcript has 2,960 words across 413 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.