Okay, welcome everyone to this Indonesia Project webinar. I'm Ross Tapsell, lecturer at the School of Culture, History and Language here at the Australian National University. It's my pleasure to introduce two fine scholars from Indonesia, Widji Yanto, who's at the Centre for Media and Democracy at the Universitas Dipo-Negoro.
And Ismail Fahmy, who's a scholar, but also an innovator in the big data field. And his company is Media Colonels, which he works with academics to produce research output on using his big data company. They're both scholars having achieved their PhDs overseas, Wijianto in the Netherlands and Ismail in the US.
And they've combined. this talk today. They'll both speak for about 20 minutes on a project on cyber troops, social media, propaganda and the future of Indonesia democracy, which has involved a number of other scholars, actually, and I'm sure Wijianto and others will introduce some of the other scholars involved in the project. But some of this work has already been published in the Inside Indonesia online publication.
So I'll put up the link there if you'd like to read more. but very interested to promote discussion and debate after both the speakers speak for about 20 minutes. So we can have time for questions where you can raise your hand and ask verbally, or you can put questions into the chat group and I can facilitate it that way. But thanks to you both, Ujjainto and Ismail, lovely to see you.
Hope you both well and look forward to hearing you present. Over to you Ujjainto. for your very kind introduction and good morning everyone in Indonesia and good salam siang for everyone in Australia.
Please let me share my screen. Could you see my screen now? Nice.
Okay, so as you could see there today it's really an honor for us, for me and Mas Ismail Fahmi to be here to present our ongoing research about Cybertrucks, social media, propaganda and the future of Indonesian democracy. I will start first and then I will give it to Mas Ismail Fahmi and then come back to me. So this is the list of content of our presentation today. Hopefully it will be enough for the time.
So it started actually with the research from Oxford University, presented for the first time, published in 2019, I suppose, in which they found that in 70 countries in the world, there were the operation of cyber troops and public opinion manipulation. And the writers were Philip Howard and Samantha Bradshaw. Okay, and they defined computational propaganda as the use of algorithms, automation, and big data to shape public life. Manual saboteur, they define as government or political party actors tasked with manipulating public opinion online. We think back then that this definition was not enough and was not explaining the real situation.
So we would like to know more. Why did we come to that town? So it... It comes back to the occasion in Capeka, the weekend of the National Commission on Corruption Eradication.
The findings of this research suggested that the capacity of Indonesian cyber troops was low, low cyber capacity, Indonesia is here, in which they only operated basically every during election. However, you know, in the time of the KPK revision, we found that the operation of Cybertruck was also there back then. Perhaps you remember, September 2019, suddenly the House of Representatives announced that they would discuss the revision of the KPK law. And the activists, including me, the academics as well, we protested it. We refused this revision because we think that it will seriously weaken the power of this CAPECA.
And so academics in Indonesia, around almost 3,000 lecturers from 41 universities from 22 provinces, we issued a... petition to refuse the revision. However, back then we got a terror, cyber terror, some of us, including me there. There was a phone call from all over the world, from foreign number, and some of us were hijacked, our WhatsApp.
And then we noticed that at the same time there was an an effort to manipulate, to make, to direct public opinion to support this KPK, this revision. Mas Ismail Fahmy will present, will share more about it later. But I would like first to explain this research as Ross suggested before, it's a project together between KITLV, Leiden, and then Amsterdam and of course LPGAS, Universitas Timor-Lagos and Universitas Islam Indonesia.
And we focus on five cases, presidential election 2019, the bill on revision of CAPECA law, new normal 2020, omnibus law and the direct relation of the local heads in 2020. Meanwhile for the resist method we combined mixed methods consisting of computational data analysis. So we basically conducted social media analysis on Twitter basically mainly and then after from the finding on the big data analysis social media unless we follow it we followed them up with field works and in-depth interview to get deep stories of motivation organizations, strategy and tactics and political networks of the cyber troops. And then we also conducted digital ethnography to analyze the narratives, cultural and political repertoires, resonance with our rejection by online publics.
Okay, and in this research we defined cyber troops consist of robot, buzzers and influencers. So basically influencers are real accounts, you know, real figures on Twitter. Buzzers are those who use a fake name.
So they use anonymous and then robot is automation. We will explain more about it later. Okay, now I will give the screen to Mas Ismail Fahmy.
Thank you very much, Mas. Good morning, everybody. And so my part now is would like to share how big data analytics methods work in this part.
Because as much we just said, we have two parts in our research. First, the quantitative method, and the second one is qualitative method. Okay, we use the system that called the Throne-Numbrate. And as a short introduction, this system actually has been and being used by Ministry of Cominfo as well in Indonesia since 2017 and to monitor negative contents like pornography and so on. And it's being used also at the Ministry of Defense to monitor any issues related to defense issues in Indonesia, especially from social media and being used also at the west Java governor sites.
So we help them to monitor the conversation about regional development and so on, something like that. And for the academic point of view, we try to help scholars using our system. So we call, we developed the Drone and Blade Academy. And this system is being used in the work together with University of Islam Indonesia. So basically everybody And here can be student, researchers, lecturers, journalists, bloggers.
As long as it's not for commercial, they can use TurnBrid Academy. So they can become members, they can propose a keyword to monitor. And the idea is that in the beginning, only me that share the analysis. So I want everybody else can help us in monitoring the system, monitoring the conversation, and they share the result for public.
either using their blog, social media, or academic writings. This is an example of a dashboard, how we monitor conversation. So many topics can be found here, and everybody that joins TwinPed Academic, they can use any topic and they can write any papers based on this. Just for example, I just take a look at the Google Scholar, then write down DroneGrid, I found almost 211 results. So at least around this number of academic papers mentioning turnip in their papers.
Okay, now we go back to the topic that Masbicha said before, how we monitor cyber troops and computational propaganda in Indonesia especially. And this one, as Masbicha said, we quite find a lot of bots and algorithms that try to manipulate opinion. Indonesia. And for just for example, beside POT we also have cyber troops. So we have JASMEF for example in Indonesia.
We have Republic Cyber Approach, 99th Army, Muslim Cyber Army, Nahdlatul Ulama Cyber Troops, okay, and Rail 1 Pride. So we can find this kind of army, especially during the presidential elections, quite a lot. So not only during the presidential election, after that as well we can find still the cyber troops. And the method that we try to collect the data actually is not only from Twitter. We mainly use data from OSINT, Open Source Intelligence, like online news, Instagram, YouTube, Facebook and Twitter.
And now we are working on TikTok as well. And from this data, we crawl almost real time, especially Twitter. So we found that for national issues or global issues, mostly they use Twitter. to disseminate their opinion, to manipulate opinion.
And while using Facebook, YouTube and Instagram, it's usually the topic rather different. And we use all sorts of technology here in a short way. Analysts, they can simply use the front end.
So from the front end, they can get any analytics that already prepared by the system. And so we have quite a lot of servers, almost 12 servers here that work almost 24 hours, seven days in a week. Okay, how we collect the data? Given one keyword, for example, or a hashtag that we want to monitor, we can just put in our system and this system can get free data in the past seven days.
So this data is provided for free by Twitter. And since then, we can have the data almost real time. So we use a streaming method provided by Twitter as well.
So there are two methods, historical method of seven days of big word and streaming methods. It is a real time. Using this method, so we can have as long as we want, as not big word, but forward. So once you have a keyword and you expect that, okay, in the next one month, probably there will be a campaign. There'll be an issue.
Okay, starting off today. We start to use the keyword, monitor that, and then you can collect the data forward. This is the methodology that we use. As you can see from our dashboard before, from the DroneBit Academy, you can see so much keyword.
And actually, the keyword, the data collected almost real-time using this streaming method. And how we can get the keyword and data? Twitter provides 400 free keywords. So for one server, we can have 400 keywords.
In our experience, in DroneOplate, we have more than 400 keywords. So in that case, we have more servers. So simply like that, so we can add more and more servers to get more data. And one of the analytics that we use is social network analytics. So actually there are quite many analysis methods, social network analysis, sentiment analysis, opinion analysis, and so on.
but in this very short time I try to explain this one, social network analysis, because this analysis is very useful for the next step and how we can detect which key influencers, which buzzers that we would like to interview and so on. Okay, in this social network analytics usually, actually we use the graph theory, very very simple, we have a subject and predict and object simply like that. like this one we have node, we have relationship and nodes. Once we have quite a lot of triple like this so we can have a nice social network analysis. For one keyword that we collect from Twitter, usually we have this kind of networks.
We have probably one big network where we have one or more influencers, top influencers, and their followers, and another influencers and followers. So in Indonesia, for example, we have pro and contra. We have a pro government and pro opposition, for example, like that. And we have other very small publics. They talk among themselves like this.
There are two interesting methods of analyzing this social network. We can use a retweet. So between two accounts, when they retweet each other, we can think that they agree.
Somebody retweet somebody else, so it means they agree with the opinion. Once we make the social network analysis, only based on the retweet, we have this kind of agreement network. And this will give us what are the.
echo chamber, where are the clusters, cluster of accounts, like this one, pro and contra. And if you mention, so people mention each other, what will happen is like a discussion, like the second graph, this discussion, actually, those influencers from pro and contra, from pro-grapherment, for example, and opposition, they will become in one network. They fight each other because they mention each other.
Okay, that's the little bit of theory about the social network. And let's see how it's happened with the KPK law revision. And we found that during one week, we have this set of keywords.
Let's say, KPK and Taliban. Actually, this one is keyword KPK, and then we will filter with Taliban because they're like an opinion building. Well, in KPK, there are Taliban. And there are a set of hashtags like this one, KPK.
patuhaturan, KPK changing and so on. And we try to find the data from the online news, Twitter, Facebook, Instagram and so on. And apparently what we can find is that most of the conversations are in Twitter. Okay, that's nice.
And the situation is like this, actually, only KPK and Taliban, the volume is huge. Within this period, I think this is about the one week period, okay, from 10 to 17. So in one week we have this huge number of conversation, Kapeka with Taliban. But for hashtags, actually the number is very small, less than 10,000.
But all of these hashtags actually appear in the trending topics almost every day in Indonesia. It appears in one or two or three hours. So you can see that with a very small number of conversation or mention using this hashtag.
we can gain a trending topic and it can attract netizens and journalists usually to give attention to what they would like to say. So this is the trend. So we can see that for example, for the on the 7th, 11th of September, we have Revisio KPK for NKRK. So it means that it's like promoting the revision of a KPK law.
It is good for Indonesia. and so on. So almost every day we have one hashtag until the end of a week like 17th of September that we know the 17th of September the law is approved.
And so one week before we have this kind of like public opinion manipulation using the hashtag. And once we use social network analysis, we plot all together the hashtag, we can see that almost all of the hashtags in one cluster like this one, the Visio Caprica for Antony, and the colors red means the sentiment is negative. So I combined together the mention and retweet.
We want to see how people retweet each other and talk to each other. And we combine also with the sentiment analysis. So in one graph, we can see the relationship and the sentiment as well. So we can see here, the red one is the negative sentiment, the green one is the positive sentiment.
So you can see that some parts are very positive. So it means that they have a very positive opinion about the capital low revision. And we can see actually this cluster is by, let's say, cyber troops or buzzers that brought the government. And this one, very small, are the cluster of activists or publics, like then the Laksono, Anita Wahid, Alisa Wahid. Those are influencers against the KPK revision.
How actually they developed this conversation? From this example, we can see... like KPK Patu Aturan is very dominant, it appears quite in the trending topic of Twitter.
So actually the conversation is nothing, it's just like a giveaway in the afternoon, one giveaway in the afternoon until number five. So they will give like 50,000 rupiah for two winners, so actually for each winner is only 25,000 rupiah. They got coupé or pulse or something for their phone okay and as long as they they retweet this hashtag and just for one uh post they get 500 retweets you can imagine like this for the second one they get 500 and so on so very easy just make for every hour for example okay this giveaway and then next hour giveaway and so on you can maintain the number of uh response and uh for that test only apparently only one account that share that and all the other account that I really tweeted.
So, you know, this is not actual account. It's like an anonymous account. This is just just I mean, I think this one is anonymous and the other one probably is like both. And when we look at this, this we analyze how the conversion works.
We can we can detect that actually the conversion is done mostly by spamming and the interacting interaction rate is very low. It's meant that. 1.94. It means that just one post they only get less than two retweets or likes. Normally it's not like that.
Normally if a human make a post they will get a lot of interaction but because it is a bot, done by bot, then bot has less interaction. So this is like indication that this very simple campaign using this hashtag is done by Anonym and That's a computational propaganda. This is an example of tweets. So we actually can easily detect how the tweets done. Like this one, we have a copy capa to run copy capa to run done by the same users.
You can see here the same hour, almost the same minutes and exactly the same boss as well. Like this one done by another users almost the same time. So they they manage the time so they can. with this way we can get a trending topic on Twitter at the time, at the same exact hour that we want.
And then the rate with as well, how people rate with that. The same account, the rate with, you can see here, that the same hour, the same minutes. It's usually done by bots or by just specific special force to do that. Okay, that's one using the hashtag. How people they use the hashtag and besides the content is only making it as trending inside the hashtag they also provide this narrative using image the image is very easy to to um to be understood by people they make quite a lot of uh graph using meme or infographic and they share together with this hashtag and they expect that once this appear on the training topic and people click that and they will get this narrative all is done by uh anonymous and maybe bought.
And they also attack the media like this one, Tempo. So Tempo Kacong KPK, as you can see in my list of hashtags before we have Tempo Kacong KPK because Tempo reports the work of these buzzers. Okay, and another way is that this kind of campaign is very difficult. difficult actually to be paralyzed, to be activists.
If they want to make it consistent, it's very difficult because you know like this example in Omnibus Law, we are only able to make a campaign, a huge campaign, only one week, let's say this one week. This Tola Omnibus Law hashtag is done by activists to then also keep poppers in Indonesia, only this period. And after that, K-poppers will go back with their daily conversation, promoting their groups of K-poppers. And after that, the world's best cybertroopers, they can use their own hashtag again and again and for quite a long time.
So we can see like we work in sprint mode very fast, okay, and then quite a lot, but in very short time. But to make manipulation of the opinion, we need not only sprint, but we need to have like a marathon, and this one cannot be achieved by normal users. So we can see here, for example, for the omnibus law, we need from October and then November and so on in December.
That's quite a long time. And once they have a demonstration, the student, they will go back to their normal activity and the rest will be done by the one buzzers that brought omnibus law. This is an example of how actually the omnibus law, it's a very short period.
Everybody work on it. we have a academician, we have a student, activists, key workers and so on, the media, we all work together and this done here, somewhere here, okay. But after that everybody is have no more power and this network is gone, okay. We have no more of this kind of network and so we can see the hashtag that being used to manipulate the opinion in the This one, we are winning here in this period, number two.
Okay, almost all of the hashtags against Omnibus Law, but after that, almost all of the hashtags pro-Omnibus Law. And it is interesting to see where are the media. So this one is not about Omnibus Law, but it's during the presidential election.
There are two clusters because we have two candidates, right? And usually during the... the conversation, we have the media, all the media are between clusters.
We have here the TIG, Narasi TV, CNN Indonesia, Tirto and everything. Usually they are in the middle. So actually we can expect that media can become like information arbitrage. So in the middle of the cluster, they can become the bridge to provide this information. So I think that's all the presentation from me and for the summary.
What I can see is that maybe in the future, the near future, the cyber army and competition propaganda will still likely be in use and they will probably have a better techniques and platform and approach. Thank you very much. Please, Mas Widja. Yes, okay, Mas Ismail.
I will continue from there. As you could see from Mas Ismail's presentation, you will be found on Twitter, on social media, the operation of cyber troops. Then the question after that was, who are they? How they work?
What's their motivation? How their organization? So to answer that question, we follow up what we found on Twitter by conducting ethnography and digital ethnography.
What we found so far was first, what is Cybertrups? So based on our ethnography for almost two years, this is actually a flip network consist of buzzers, coordinators, influencers, content creators, and political consultants who work together to orchestrate public opinion on social media by creating... a certain narrative on certain political issues in which every change of the network don't necessarily know each other nor they necessarily know that they work together for supporting the narratives that's why it's a little bit complex and fluid and who are they we found that they were mostly young based on our interview with 78 informants well-educated men so only small number of women there below 45 years old, generally 25 to 35 years old, well educated, mostly at least bachelor degree, computer saviour, creativity, digital literacy, they have social media militancy as well, motivated both ideologically and or commercially, working voluntarily or for payment, and the metamorphosis was They were initially volunteers and then became project-based buzzers and then coordinators. Some of them got high status in government or business or owner of a buzzer company. Some of the cyber troopers, for instance, when they got married, the minister came to their attending their marriage to become the witness of the marriage, Saksinika.
Okay, and we funded the funding of... and organizations buzzers more or less they got 50 until 100 000 rupees per account or 207 million per project so buzzer those who use anonymous accounts yeah to to bust something and can content creators they got more or less four million per project coordinators they got 200 000 per account or 5 to 15 million per project and for influencers they got around 20 million rupees per project or political position as commissioner or government project. So one disclaimer here that out of 78 informants, not all of them of course share this thing and the salary or this funding, but for those who answer our questions, that is what we got. And where are the resources of the...
the funding. First of all, we found that this is an individual politician. Usually from, so what the government suggested, this is usually from a politician's personality. Legislative candidates during the presidential election, they have lots of money. And in the case of common support, raising the name of a figure to be appointed to enter the upcoming election markets, they got paid only for comments.
on his Twitter account to make it more famous. So, you know, in a time like this, as we are doing research right now, some of our researchers, because they were here before, they were social media activists as well, they got offered, for instance, to get boosted, to boost one politician for the 2024 election. Okay, and It's also from political parties. So one former suggested the source of the funds came from elements of political parties, some from elements of donations. So that's how it is.
Those who donate are the ones who have the money. Those who don't have money can contribute energy and thoughts. And also from government, and we found it was mentioned quite often in the case of normalizing new normal. So one informant suggested, yes, there are several. The ministry asked for it.
We get it from the ministry. It's part of a socialization fund. This is for TV. Not always TV.
You see, you don't have to go to the media. So I have a projection that later, the budget for socialization public relations will be divided into three, social media, online media, and TV. Okay, and this is the emphasis for the new normal.
The new normal is different. New normal is Jacob's interest. There's a minister in charge. There are three ministers, ministry in charge.
Well, maybe the funds come from there, the budget. So this one buzzer suggested to us. And also from the business person. So during the election, this was the case in the 2019 election.
So many... in the trainers or business person approaching the candidates one of them was approaching the vice presidential candidate and our informant who was the close uh person uh in the in a vice presidential campaigning team uh uh he suggested that uh some of the businessmen were rejected but some of them were accepted i mean the the money okay so what uh how was the languages of the social media propaganda We found that in our digital enography for this part, this is the kind of language. So, attacking those who are critical to the governments are some of the language. The rumble is based on the government, the lockdown, and the praise of Vietnam, the communist country. The COVID-19 pandemic, the coronavirus, is a new face of the PKI government, for instance.
So, they are attacking them as communists. those who are critical to the government or social justice warrior are named for those who are critical and this is for the case of other case yeah in papua uh this is not cyber troop this is i think dr ross apsel yeah promoting our seminar this morning uh on his twitter okay another language is using uh doxing you know uh so taking a personal account of someone to put it on twitter so uh This is for instance, in the case of SDM, students who protested, they took the screen capture of the conversation and put it online. Okay, and this is not the method of doxing.
So in addition to the five cases that I mentioned, I also conducted, as you could see, you could read inside Indonesia. about the digital coup inside the party democrats. So this is the kind of thing.
So Ahalyay back then was phrased as having an affair with someone else. So they cropped the picture Ahalyay with a woman and then they frame it as you know you know as a China terlarang or as happened as the chairperson of a democratic party, a party democrat having a mistress. Okay, another way of doing it is trolling. So, you know, trolling basically is sharing content, hostile content to start initiating a quarrel.
And this is, I hope Mas Ismail doesn't mind, yeah. This is the example how Mas Ismail Fahmy got got trolling for his posting. I think he got it a lot because he's quite active on Twitter, while I'm not so active, so I don't have this kind of things there.
So yeah, basically they attack you personally and then they initiated the war or fight there. Okay, and as Mas Ismail Fahmy mentioned before, attacking the media is also one of the of the trailing method of the buzzers. So to reflect before we close, to reflect about it, so this is actually a reflection Dr. Ross Tapsell here and I'm Xin Peng wrote a very interesting book about from grassroot activism to disinformation. So this is a trend actually in Southeast Asia as well as in Indonesia. In the conclusion of the book Marco Bunte wrote a nice lines in which this is And the trend was the internet today, the social media today was used by the authoritarian regime to become authoritarian.
And the newly democracy, the regime used it and as a result, we have a democratic regression. And which is reflected also in the publication of LPTKS, for instance, that we have a democracy tan pademos, reflection of, you know. more than 100 scholars from all over the world, but in Indonesia we now have a serious democratic decline.
But in this occasion, I would like to go further, that this is also actually a reflection of the rise of digital authoritarianism, in the use of different digital technologies for surveilling, suppressing and manipulating citizens. using the dissemination and misinformation and so forth. Okay, so the conclusion before I end it. First of all, there has been a manipulation of public opinion by using propaganda on social media to support problematic policies that have the following characteristics. So there was a sudden tsunami or tweets with narratives supporting certain political narratives.
So content creation and intentionally and by a team of professionals to support the narrative. So the content was made by professionals, not just you know just iseng-iseng or just not serious matter. And there's a usage of giveaway to attract organic Twitter as well.
The narrative at first gets a match or you know or challenged by the civil side but in the end because they win the marathon like Marcy Smiles suggested. So they have the endurance and quantity and duration. And this narrative was amplified finally by the mainstream media. And during our workshop, I could add more that the use of digital primalism is one of the characteristics of Indonesian cyber troops. The second, the manipulation was conducted by a cyber missionary of networks of buzzers, coordinators, influencers, content creators, and political consultants.
who work together to orchestrate public opinion on social media by creating a certain narrative on certain political issues. The funding of this missionary came from individual politicians, political parties, government, and business persons. Third, at the end, the operation of cyber missionaries and the public opinion manipulation they have conducted was aimed to manufacture constant diplomatic policy through among other things spreading hoaxes, fake news, doxing, and trolling.
Fourth, this in turn has been successfully co-opting the Indonesian digital public sphere and has been preventing it from becoming a free public sphere where the voice of the civil society can be heard. In this regard, the online public opinion manipulation by the cyber-ministries can be seen as one of the most significant signs of the rise of digital authoritarianism in the fourth largest country in the world. And this is the look of our publication in Indonesia. Thank you. Wonderful.
Thank you very much, Widja and Ismail, for that really comprehensive explanation of your project. As Nuka has said in the chat there, if you would like to raise your hand, you can raise your hand and ask a question. And I can call on you directly to ask that question or if you'd like to type it into the chat room, you are welcome to type it into the chat.
Obviously, in topics like this, you know, there's no such thing as a dumb question. Feel free to ask technical questions. I've spent hours with Ismail in Jakarta and Zoom, and I still don't follow everything that happens with his methodology.
So feel free to ask, but of course, do try and keep your questions relatively short, not long treaties or long, you know, monologues. So the first question, hand up, I think I've got this right, is Ed Aspinall. Hi there. Can you hear me? Yes.
All right. Thank you. Thank you to our two guests for a great presentation. Very interesting.
So I guess my question is sort of a bit of an obvious one. So obviously we see these buzzers and your presentation is really very thorough in outlining, you know, how they're organised and what they do. And we see they're very successful in flooding social media with this material.
But I guess the question is, what's the effect beyond that? Do we know if they really do have a bigger impact on public opinion, on shifting media narratives? It might well be, I mean, what's the ultimate purpose of just flooding social media with all this material, which, as Ismail shows, gets very little interaction. It's just sort of sitting in there as a sort of dead weight.
in cyberspace but perhaps not really influencing people. So I guess that's my question. All this money which is being spent on this, is it really worth it?
Is it really affecting how people view political issues out there in the, I shouldn't use this term, but out there in the real world? Thanks. Yes, silaka. Ouija. or if not go answer straight away yeah oh okay um thank you um uh at uh good to see you again um so is the world it it does it in impact the media or the public opinion the answer is yes um they will not spend so much if it doesn't have the impact of course I would like to show you something here that I prepared.
So I will start with this, this KPK case. Yeah, this is the coverage of Kompas Daily newspaper in 16 September 2019. So, and the KPK, the new KPK law was passed 7th of September 2019. So, this was the survey a day before the law, the revision was passed. And the public opinion back then, you know, every year, every survey institution, lembaga survey, they always announced that KPK was the among our... the only institution that is trusted the most by the citizens based on the survey from time to time.
However, suddenly, the result of the Pulse survey of Kompas, 44.9% or the majority of the Indonesian public agreed with the revision of the law, the KPK law, while it has great criticism from the public. So our suspicion was that the public opinion manipulation on the internet was affected this. But why we come to this?
Of course, this is a hypothetical question. Why do we come to this? Because we found that 250 articles in Indonesian cover we found in our big data analysis reporting the frame of the cyber troops that KPK is Taliban, you know, KPK and Taliban, KPK as the place of the Islamic hardliners. And it's amplified by 250 articles in mainstream media.
If you got your voice on mainstream media alone, it's already something. Because I'm a student of journalism as well, so I learned there that in the theory of eigenasetting, the mainstream media may not always be successful to get you to agree on one thing. For instance, maybe the public...
do not agree that back then the KPK was the place of Taliban. However, at least they will think about it. And the propaganda theory suggested that fake news, if we got it every day, slowly we will find that maybe it's right or maybe it's even correct. That's the doctrine.
if every day you got the news that KPK is Sarang Taliban, KPK is the place of hardliners, slowly I found back then that some of the publics, they begin to think, oh maybe that's right, maybe that's not entirely wrong. And based on our FGD, Focused Group Discussion, with journalists and with some activists, there was Mbak Anita Wahid who suggested that she believes, she got news that even the people in the presidential palace, you know, in the Istana, they also believe, and perhaps the president himself, he believes that KPK is the place of hardliners. So the impact of this information not only impacted the public in general, but also the elites, the decision makers.
They believe, they got to agree, and they start to think that oh if this is the place of hardliners and in indonesia if we if you use political identity like islam liners it's very very powerful so that's maybe one of the cases in which i could give you illustration about how powerful it is uh the cyber troops and the public immigration they did done uh not just impacting the public but also the political elites that would be my point maybe you must smile have something yes i just want to add a little bit So not only during the KPK, just a few weeks ago for example, in Indonesia we have a group of buzzers, they say, Bu Barkan MUI, MUI, Majlis Ulama Indonesia. This actually is done by a very small number of buzzers, but it appears in the trending topic. What happens then, the journalists, not only the online media but also television, they found this very intriguing, they found this very very, very, very catchy if they put this in their program.
And since then, few days, almost several days, they make this program. They interview one and another and then they try to interview me as well because I'm part of the MOE. I cannot become part of the Nara Sumpar there. And so this is an example that actually is very interesting.
You can make a very. small voice, but you can make it this trending. Once you can, you can make a very nice narrative, and this narrative is very interesting and intriguing. And people will catch up and especially if the media catch up, then boom, it will come viral. So that's the way how the buzzers and manipulation, public opinion manipulation is done in social media.
It's not only in social media alone. we need this to be viral and we need this to be viral on online media, mainstream media. And once it goes to television, then you're successful. Right. Yeah.
Thanks. Which would be, I guess, the standard line people would ask or would answer Ismail, if people say what percentage of Indonesians are on Twitter, right? It's like not, not even 25%. So, so you're not looking at huge engagement, but the idea is Twitter is an agenda setter that then hits mainstream media content.
Yeah. you know interesting um let's go to marcus he's got his hand up and then i'll get through uh these q and a's over you marcus okay thank you i want to start my question with a clarification question to uh vidya because what he just said confused me um You at the end sort of portrayed the palace and Jacobi as a victim of this misinformation, if I understood you correctly. But if I follow your presentation prior to that, and that made more sense to me, then it's the power holders, including the palace, that is actually the source of that misinformation.
And it trickles down through a layer of middlemen that... then hire these influencers, buzzers and so forth to spread a narrative of a political subject that is sympathetic to the government's line. So can you clarify, you know, given what you presented to us, can you then clarify what you meant that at the end of the day it's the political decision makers like Jacobi who are also the victims of this when in fact it probably is the other way around.
That leads me to another issue I wanted to ask about, and that's your reference to cyber authoritarianism. Was it online authoritarianism or cyber authoritarianism, which is an interesting concept. But going back to what Ed was saying, basically what you have here is the government flooding the market with positive stories about themselves, that in itself is not an authoritarian measure. However, if I go back to one of your graphics there, which was interesting, that showed that after the arrest of the Kami leaders, the anti-government expression of opinion declined, then I think you do have a better case, but you have to really connect the two. So my question on this particular case would be, what is your evidence that the arrest of the Kami leaders led to a decline in the expression of anti-government opinion on the internet?
Or do you simply connect the two based on the date you have of the arrest and the consequent decline of... anti-government expression because that's the sort of thing I really you would really have to go into to make the case for authoritarianism rather than an increase and the flooding of The market with pro-government propaganda. Okay, Marcus before I answer your question a bit clarification to your question Comey leader Could you elaborate more about it? That was in one of the graphs that Ismail showed us. So maybe Ismail you can explain to him what was in that graph.
Yes, actually like this Mas Widja, you know we have demonstration in the first week, everybody's going on like that. Then in the second week, Kami leaders is like there's a group of organization Kami, you know. So we have two leaders now is being captured Shah Ganda.
Okay, and Jumho Ridayat. Okay. And this also gives an impact probably.
So the question is that whether the capture of the kami leader gives an impact of the decline of the voices from the anti-government. And this is Ketut Normantir's kami, right? Yeah, yes.
Okay. Yeah, thank you. Yes, thank you, Ed, for your questions.
I think we learned from your questions. First, about why I suggest that the president might be the victim of this manipulation about KPK as the Sarang of Taliban. Because based on our research, we found that it was from the palace, yes. But in the palace, there are a lot of powers there, right?
A lot of people, a lot of actors. So it's just not... single power. And we found, maybe we can link our research, at least, we found that some people in the palace, we suppose they were involved in it. However, we don't find, we have to be honest, we don't find at least explicit evidence, for instance, that the president himself directed or instructed.
So in a way, so we maybe we learn, we learn, at least myself, we learn that this is people on the palace want to influence the president himself. want to speak to the president, Mambisiki president, and they use as well the cybercoups to convince the president. So that will be one of my thoughts.
I hope... Please give your comments if this sounds confusing for you or not, or maybe you have another way to think about it. And... That would be the first answer to the first question.
And the second question, I disagree. It's not just flooding the social media with the news about supporting the government, but because there were also misinformation used there. For instance, in the case of KPK, that's obvious.
It's not just that. they don't only support the revision, but also they use misinformation that KPK is the place of the Taliban, and that's not right. And also in the case of weakening the Democratic Party, I think I presented as well in this you were there, about how the Partai Demokrat as an opposition, they got attacked in social media as well. They got misinformation about their leader, Ahaye, that I have presented before, and also some others systematically to manipulate public opinion to support the KLP, the Congress Warbiasa of the Democratic Party.
So they weakened the opposition. One of the ways to weaken the opposition was by using the social media to get support from public opinion to make this the party democrat weak. So weakening the opposition I think is one of the attacking the opposition, one of the signs of authoritarian regime. I agree with you as well.
One last case about the kidnap of Kami. So what they used in the case of Kami leader was accusing them by, you know, they did, they did the hijacking the WhatsApp group, the conversation of the community, and based on the result of the hijack of the conversation on the WhatsApp group, they used this evidence to send the person on jail. So it's a combination, I think, of public manipulation and cyber terror in that case. So it's beyond this study.
But I also had study in other case about the hijack and also cyber terror about in the Kabeca case. Thanks Wija. We might try and move on to the people who've typed questions now.
Narva, did you want to ask your question or shall I just read something? I'm not sure if you're there. Narva's asking about MCA, pro-Islamist buzzers.
I know you've been working on this for a long time, Ismail, and looking at some of that. material and data. Do you find different comparisons between the pro-government buzzers and the pro-MCA?
Do you want to start with this one Ismail? Okay, so the difference between MCA, Muslim Zebra Army and pro-government buzzers. MCA is mostly voluntary and when I look at during the residential election, MCA become one of the supporters of Prabowo for example. And when I look at the way they create a narrative, the way they communicate and spread the narrative is very different from the group that's supporting the government. The group supporting government is very organized, so they have a narrative, they have goals.
Okay, for example, this week what are the narratives that they would like to share, positive things or negative things. Positive things to promote a positive program from the government, for example, or negative things that attack the opponent. And the next week they have another narrative and so on.
And they can provide materials, for example, using memes. video, or articles, all organized. It's very nice. So we can see this almost every time.
You can look at this as well in my presentation before, how they use the meme, how they use the infographic, it's very very nice. But for the other way around, the opposition or Muslim separate army, and now they say that opposition, they're less organized, they don't have a narrative, they don't have organized narrative. they're usually like sporadic.
Okay, now we have to, we have this hashtag. Let's pick this file. The next day, they have this hashtag, next file.
And during the presidential election, they only have one hashtag that was used for a long time. 2019 since the president. Only that all the time. And it's very different from the buzzers that is supporting Jokowi.
Almost every week they have their own narrative, different hashtag. You know, making hashtag, making narrative, it needs a group of people and make a team. They have to analyze what are the concepts, what are the situations, what are the situations of the social media, and so on and so on.
So they can make a good and logical narrative in this way. But I cannot find this on the Muslim Seperate Army or the opposition. One is more very organized and the other one is, I can see, less organized. Great, thank you. Anshley Bexley has a question here asking about the gendered nature of the fact that so many men are doing this work, therefore are the narratives gendered?
So, for example, you used the example about, you know, having an affair and these kind of things. So are you seeing a gender relationship between the workers and the narratives? Anyone?
Widja, do you want to start? Yes. I think the short answer to that may be yes, because mostly the social media activists or the cyber troopers are men, so it's naturally the view is male dominated perspective.
However, we don't analyze further about it. So thank you for the question. Is Mal anything to add or we'll move on? I think we can move to the next question. Okay, thanks.
Yeah, great question there. And something probably should be, more work needs to be done on that. Avi's got a question about the Undang Undang Ete Air, which I presume none of the buzzers have been arrested under Undang Undang Ete Air, and it's mostly activists and journalists.
But does anyone have an answer to that question? Is there any way that the law is affecting buzzers? Um, so...
the law may be affecting buses but from the opposition not from the pro-government uh we have uh so when you say opposition buzz do you mean like cracking down on that on the muslim cyber army for example uh that that kind of thing um we found in our research uh the supporter of prabhu during the 2019 election you know uh that day uh and um he he felt he quitted from cyber trooping. He said, because you know, Mas Widja, if I, you know, because I'm pro-opposition, it's very vulnerable for me. My friends who supported, my friends, they also buzzer, who supported government, they're fine, they're just fine, they're okay, they can continue their works, but not me. So that would be one of the example of the situation.
Great, thanks Widja. Jordan's allegedly got two quick questions but they look pretty pretty long jordan but i'll do my best do you want to try and uh jump in here or does anyone want to ask the question about can you provide further comments views on buzzers activities activities targeting islam's opposition activists so you sort of just answered that one um anyone to add on that on the campaigns against the car omnibus law And can social media analytics tell us anything about genuine public opinion or is only useful now for understanding politicians, activists, online campaigns? Yeah, so that's somewhat similar to Ed.
But it is an interesting point. Like anybody now that's saying I've got all this Twitter data for you and it's representative of public opinion. I mean, you guys have shown that it's hugely manipulated by buzzers. Does that mean, therefore, that any...
social media company like Polytweet or, sorry, that's the Malaysian one, or what's some of the Politica Wave, these kind of organizations that say, I've measured sentiment on social media, including Twitter. It's got to be dismissed. What do you think Ismail? You would still advocate that you can gauge sentiment on social media wouldn't you?
Yes, when we look at this question about how buzzers activists targeting Islamists or whatever, when I look at the way they use the same term all over again like for example Qatrun okay Qatrun and then Taliban for KPK Qatrun. for depicting the opposition or everybody that against government, they said Katrun. And so it means that there's like the war of definition.
They define the words and they use the words to attack others. And they use the words all over and over again and again for a quite long time. And this will be very effective in making opinion of the public that Those who are opposing the government are cartoon. And you know, they define cartoon as somebody less educated and just one-sided and something like that. It's very negative.
So it's very easy. So next time, if you want to attack your opponent, you just say that one word. That's enough.
Because you already have invented that word for a long time. And I found this very practical, very smart. And the most important is finding the words and finding the definition that you want to use and to define others. And just like now, for example, lately because in MOE, Majelis Ulama Indonesia as well, we have Tuparkan MOE, and then they try to define MOE as a cultural as well. Because you know, there's one of the terrorists suspect the terrorist as part of the MUE.
And then they say that the whole MUE is like, seems like a lady infiltrated by a terrorist. And then they use that word again and again. So they use that word MUE and terrorist together all over again and again. So this kind of way, as you mentioned as well, it's like a marathon. You just do it almost every day with small number of people that speak on it, small number of buzzers.
But as long as you can do that all the time, you can get an impact on the public opinion. I would like to add a little bit for that. So if I understand the question correctly, it's about Islam and political identity, I think.
So, you know, I did a research on corruption or kapeka since my master degree. So it's been years. And I noticed that KPK remained strong.
From time to time they got attacked. So we have Tichak Buaya, Gieco vs Cradle, series one, series two, series three and so on. And many works on this like the one from Fiona Suwana and even Ross Tapsell in his book Media Power. I think he addressed the or somewhere else.
Yeah. correct me if I'm wrong, about the power of the public support for the KPK, as well in the media social as well, that KPK remains strong. But in 2019, that didn't work. There was no public support. At least, you know, the public opinion supported the revision of the KPK law.
What explains this? That's something that puzzles me. And I think, so I learned about what difference, how did the attack of KPK differ in this time?
And this is the use of Taliban, Taliban sentiment, KPK versus Taliban sentiment. So in my view, it means that I agree with Mas Fahmi, Mas Ismail, about the powerful role of identity politics in Indonesia. You may be use anything, but as long as you don't use the political identity, framing someone as Islamic hardliners or even Taliban, it will not be very effective. But once the word Taliban is used, everyone understands, everyone discuss about it.
Some people in KPK are busy. countering the narrative. There are some Christians as well in KPK.
How could it be that you say that in KPK is Arang Taliban? But even the counter argument means that the attack is really matter. that people really discuss about it and it does the trick, the Islamic headliner frame, Taliban frame.
Okay, thank you. The next question is Tito. I hope you're enjoying your lunch.
Asking about YouTubers. Anyone got comments on YouTube? the role that YouTube, which is often underrepresented, right, in the studies of social media disinformation and buzzers.
It's often Twitter data and we don't look much at YouTube. Ismail, to start maybe, do you think that YouTube plays a significant role in creating a lot of this fake narratives and so on? Which question is that?
It was from Tito Ambio. I think it's just gone. They might... Oh no, it's still there.
Do YouTubers play a role? If yes, how? If no, why? Okay, I can answer that.
In my experience, when I look at data, YouTubers play less role in the conversation in public opinion manipulation in this case. Mostly Twitter. Because it's very easy. You need only one day to be... less than one day and you need to be able to make your content disseminated very fast and very short and very easy through WhatsApp.
So actually when we monitor Twitter, it's actually in my opinion, in my mind, it's not only Twitter. It is like the top of an iceberg actually. And what is being sent through Twitter is like the narrative that they send as well through WhatsApp and other channels.
Of course, they're farmed, suitable, freshable on TikTok, on YouTube, and so on. And yes, YouTube will become a very useful example, but it takes more time and effort for you to make a nice video. The most effective is if you can make your video very short, one minute, or one or two and 20 second minutes, you can share it through Twitter.
that's the most effective. Sure, thanks. That's a great answer.
You use YouTube for example, Chokro TV, they use that channel as well to make a lot of followers and they use YouTube to give a more thorough and more comprehensive idea. For example, if they have narrative that they put in long form in the YouTube but then use that one and cut one by one and make it short and then share through other social media. Great, thanks. Widja, would you like to answer very quickly, I guess, because we've got to get through a few, but that Indonesian journalism, the biggest finding really is that the most influence is actually the way that the mainstream media pick up Twitter and Facebook content and then make it out to seem like real news.
What would be the lessons for Indonesian journalism, Widja? Yes, I have explained the finding to many Indonesian journalists from time to time, and I think they started to get the idea and they shared an agreement with me that, you know, maybe the journalists also contribute to the amplification of the fake news, of the agenda of the cyber troops. like in the case of KPK.
Even if the news is denying the insistence of Taliban in KPK, but still in terms of propaganda theory, they take the source of the cyber troops to follow this up to become their coverage. So we suggest that if you want to clarify that there's no Taliban in KPK, so use other words, don't take the word Taliban. because it will influence the people.
And we found it in almost every case that the cyber troops agenda is finally amplified in the case of KPK, Sarangkaliban, Unormal, Omnibus as well. So we think that the journalists should do more role. as a clearinghouse of information.
However, I think, Ross, you also know about it, that the ecosystem of our media now is clicked by journalism. They do for writing, they want to get the news quick, they want to produce news that has a controversy that will attract the readers. And most of our journalism are still like that, with some exceptions. But even the mainstream media, reputable mainstream media like Kompas, as you could see, they also take it seriously, the case of KPK, and they do all about it. And one day before the law was passed, KPK reported that most of the public supported the revision.
So, you know... We could say why Kompas for instance not decide not deciding to not cover you know to not report to get it public yeah they can they can do that actually you know because in the case of Prabowo versus Jokowi in 2014 based on my dissertation research I found that Kompas decided not to publish you know the fact that Prabowo back then was a little bit ahead of Jokowi. So they could do that.
Great, thanks. This is a really good next question. What is your opinion on dodgy online media Twitter accounts such as Galora, Opposici Cerdas, Demokrazy News, etc.? This is from Ario. I often see them being retweeted by opposition figures like Fadli Zon and Rizal Ramli.
Are they also considered cyber troops? I know there used to be C-word pro-Jokowi but I don't see them too active anymore. Maybe Ismail?
Yes, yeah I consider them as a cyber troop and part of the troop okay and we have C-word on the other side, we have a Jokro and from the opposition side we have and of course this kind of cyber troop can be very well organized and paid by somebody or voluntarily organized or something. Galora Oppositores and Democracy News is very effective in terms of expressing the narrative from the opposition. Mostly the opposition will use this kind of media from themselves to share it because they can make very nice and clickbait titles. according to what they want.
You're still muted, Ross. Sorry, classic error. Let's answer Gita's second question here.
Currently, the public debate is on the anti-sexual violence bill, but somehow the buzzer influenced a machine from either parties were not as aggressive as back in the Kapeka omnibus law period. is that what is your opinion on this yes um you can see the questions right guys yes yes i can okay uh thank you again and um i think from from the drone data we we cannot see which one that they would like to docks but uh this really depends on the narrative about the issues, depending on the issues. If the issue is very sensitive, they thought it's very important, we can take a look at the data from the drone, for example, so we can see which one is very influential, that can be dangerous for that issue, and probably they will get attacks. Yes, we can do something like that, but actually, not all the issues, we can expect that we can have toxin.
It really depends on the issue, whether the issue is very sensitive and they would like to keep the issue safe for their agenda. I think that's one. And I think for the second question, like, related to anti-sexual violence bill, why the buzzers from broad science is not as aggressive as when we talk about the KPK and the Oblivious Law.
Probably the. again it's about the topic about the issue the anti-sexual is maybe less become attention of them and if we look at maybe the pleasures mostly maybe men or something like that they will see probably i don't know maybe you can we can talk more about this on the social and political point of view but when i look at this and uh the anti-sexual feeling feels when i look at the data it's the the number of confession is very low and quite Other bills, for example PDP, Perlindungan Data Pribadi, it's very low as well. When I looked at this bill quite a long time before, I wonder why public have no attention, very small attention on this bill. But apparently this bill is very important for them because they don't understand about the bill. And in terms of CAPECA and omnibus law, it's very clear, I think, about the CAPECA.
The public already know about the CAPECA issue a long time before, and they have supported CAPECA a long time before. And CAPECA is something that almost every Friday we expect that the nice news from Caprica, you know. So actually the issue of Caprica is really very popular and people like to join on that as well.
And omnibus law in the beginning, very small people work on this. If you look at this data, during the period the government defining the omnibus law, talking about this and so on, very small people. they talk about omnibus law only activists, journalists and academicians, not many.
But once we have demonstration, you know, if we have no demonstration, then the conversation is very long. But because we have demonstration, then people start to talk about that. And the key poppers, you know, they look at this because we have demonstration. If not, maybe they will not support this conversation.
So I think quite many situations where when this issue can be very high, the other issue is not. Great, thank you. We might move on, we got a few more, so we might try and get through a few more in the last five minutes or so. Tom is asking, and I'll go to you Wije on this one, as long I won't read the whole thing, but do you think these dynamics will continue towards the general election of 2024?
and we'll see more sophisticated tactics from both camps, pro-government narrative cyber troops versus civil society. Widju, do you want to start on that? Yes.
So that's what I would like to add as well to Masi's questions. Yeah, before, I mean, question to Masi's smile. So what we learned from the cyber troops, you know, the metamorphosis of the cyber troops as well, itself, yeah, from...
in the first as voluntary and then it becomes commercial it is uh politically motivated right so uh it becomes more and more sophisticated and it also becomes more and more professionals quote unquote it means professional means they work based on payment now that's what one of informants our informants got uh and it will be uh continuously used in the future so one of our uh conclusion from the research and also from dissemination is that, you know, the phrase yang fana adalah waktu, pasukan siber abadi. So maybe time will change, but the cyber troops will still be there in our Indonesian politics and shape our democracy, and it will be increasingly more relevant in the future. They will play more role in the future.
In the old days, people will win the election only by food buying, money politics. However, now it's not enough. They have to do the digital campaign as well and it involves the use of misinformation very often, political identity as well, so it uses cyber troops. And to add some points to the previous questions, why some issues are viral while others are not, I think it has something to do with the oligarchy.
Issues like KPK, Omnibus Law, New Normal, it can involve a lot of money there. So they can do the marathon for the campaign. They can do the viral because there's money in it. There's an oligarchy who have the money to fuel this. But for some other issues, maybe it's not their interest.
That would be our thought about it when we talk about money and fundings and where they're from. and what's the interest behind it. Okay, thank you. Ismail, the next question was about social network analysis is largely urban based and middle class. Do you disagree?
Can you share findings to prove me wrong? That would largely be the case, right? The tweets that you're looking at would be citizens who have internet access in urban areas, right?
Okay. So what I would like to say in this case is that, you know, in the very beginning, we always said cyber truth and computational propaganda has a purpose in manipulating public opinion. So it means that it is not representing public opinion, but to manipulate. What does it mean?
We don't have that opinion. at hand before. We don't have that opinion whether it is in an elite, the medium or in a rural area. We don't have that opinion but we want to have an opinion. We want to have a narrative.
Just for example, KPK is Taliban. There's Taliban in KPK whose opinion have this. Nobody have that. But using cyber truth and computational propaganda and social media, they try to create that opinion.
So this is what we would like to monitor almost during this opinion wars. We want to see which kind of opinion that's currently being spread to manipulate public opinion. And it works.
For example, Taliban, the issue, we have monitored long before, when the very first time it appeared, and not many people believe on that, but because they use this word of all again and again and again, then quite many people will have that opinion. So that's the goal of this Cybertruck and computational work. And I agree if you say that SNA is not representing the whole population and maybe in a rural area that they have an internet connection.
No, that's not what we would like to see as well. Okay, and maybe last one then from Adi is asking about how difficult it was to independently link the buzzers to their funders because obviously political parties are obligated by law to disclose their campaign expenditures, but they wouldn't in the case of buzzing. But how hard was it, Widja, to independently link buzzers to political parties in your research?
How did you do that? Yeah, actually it can be very difficult, very tough if you start from zero. However, we were lucky that our teams, they were retired cyber troopers, some of them, and some of them maybe retired for a while, I'm not sure.
But because they're insiders, so we can go to the sources quite easily. But still, having said that, what we got on the field was that we suspected, my impression was. and I know this person from political party, for instance. So of course we will not find any evidences, let alone legal evidence, that there's money for the cyber troopers. So we have to...
This is an underground politics, underground campaign, and everything was in a dark mode. And not every informants are willing to share their information about the findings, for instance. So that's what we have learned based on this almost two years research.
Great. Thank you. Well, I think we're out of time, but thanks again to Wijan and Ismail for a fascinating presentation.
Thanks to the Indonesia Project for organising that. If you could all give your digital applause, that would be much appreciated. And thanks, there's been a poll to go around as well.
Are you looking to share those results, Nukka, before we wrap up? Or is there something more you wanted to do? Yeah, just that's the poll. I just want to announce for a global seminar next week, if that's okay at all.
Sure, sure. Next week, we have a GESI global seminar, the first one, by Dr. Angie Bessley from ANU. And she will be talking about Ta'aruf online.
new spaces for female agency in indonesian digital instagram matchmaking so quite similar will be very interesting okay thanks wonderful well thanks again to you both and for wonderful research and um thanks to everyone for attending we'll see you next time thank you ross thank you everyone thank you ross thank you everyone thank you masuija thank you thank you man thank you um i'm going to close this now yes yes okay thank you bye