The Making of an Anti-biometric Mass Surveillance Campaign

By Filip Milošević, SHARE Foundation

In Short: This case study describes a series of actions carried out by a community of activists during an 18-month campaign against a new mass surveillance project in the city of Belgrade, Serbia. Not only did they manage to raise awareness about the problem, but they also engaged citizens in joining the movement. After the campaign, the government finally admitted to its lawless tendencies and halted the facial recognition software rollout.

Even though this text was written by a single person, it is a story of many, narrated from the perspective of several digital rights activists from the SHARE Foundation who coordinated the campaign.

The SHARE Foundation is a digital rights organization from Belgrade, Serbia, which gathers researchers, engineers, lawyers, activists and artists to better understand emerging technologies, strive for open internet, and empower fellow citizens to counter mass surveillance by governments and corporations.

How it started

One cold winter evening in Belgrade, Serbia, in January 2019, there was a rather odd public announcement made on the Serbian national TV:

“There will be no significant streets, entrances or passages between buildings that will not be covered by cameras. We will know from which entrance and building the perpetrator came, from which car…” (author’s translation from Serbian.)

These were the words of the Serbian Minister of Interior (the man in charge of the police) at that time, who boasted about a new smart surveillance system that our government had just bought from the Chinese company Huawei. He said that deploying 8,000 smart cameras would make our capital city of Belgrade safer, and even help fight organised crime and terrorism.

As researchers at the Serbia-based SHARE Foundation - with a special interest in surveillance - we were shocked, but another thought instantly came to our minds:

“Many of our fellow citizens actually see this as a good thing. Who wouldn’t want more security and less crime?”

As you read this, there is one important fact to know about Belgrade’s street safety record. At the community- and street-level, Belgrade is a fairly safe place for citizens. There are however documented high-level interactions between some of the governing actors and organized crime, which go far beyond the street-level safety that 8,000 cameras would be able to guard.

After this public announcement, many questions started popping up in our minds:

  • What is the real purpose of this new street surveillance project?

  • Why was there no previous public discussion?

  • Is such a mass surveillance system even legal in our country?

  • How will it affect our society?

  • And finally, how do we - the people who care - stop this?

Some details about the situation in Serbia, as of 2021-2022:

Lately, Serbia has been considered as a hybrid regime, officially a democracy striving to join the European Union, but in reality, more of an authoritarian system under strong influence from China. The government often controls the media, threatens journalists, suppresses freedom of speech and protest culture, and smears independent investigative journalists and political opponents.

Obviously, covering the whole city with cameras that recognise people without their knowledge and consent will help with this kind of social control, producing a strong chilling effect as in panopticon prisons.

  • “Those who already have disproportionate amounts of power stand to gain more and more, and those that are already in positions of powerlessness will be made even more powerless by this dynamic of who gets to watch and who is watched.” (quote from Ella Jakubowska from European Digital Rights (EDRi) in a conversation with SHARE Foundation’s Filip Milošević, full interview here.

We realized that we needed to understand and expose the real intentions of the deal with Huawei, to help citizens understand the consequences of this system for our society, and to offer them ways to join an initial campaign as well as a longer-term movement so that together we could create a counter-narrative in the public discourse to oppose the false narrative that the government promoted.

Clarifying some key concepts used in this article:

  • A false narrative is a story that is perceived as being true but has little basis in reality.

  • Political narrative refers to storytelling techniques and discourse used by politicians in order to present a topic or a problem from a specific angle. It aims to influence the attention or action of the audience (citizens, voters, etc.) by offering a certain perspective of the reality (sometimes in line with, or contrary to facts), which may serve particular political interests.

  • Counter-narratives are ideas, discourses or stories that emerge from marginalised, non-mainstream groups as opposed to those created and promoted by dominant and powerful actors promoting what can be considered a mainstream or a “master” narrative. The concept of counter-narrative implies reaction, resistance and opposition (sources: “Counter-Narrative”, paper by Raúl Alberto Mora and “Routledge Handbook of Counter-Narratives” 2020.)

Initial research and the first steps of the campaign

At the SHARE Foundation, we immediately sent FOI (Freedom of Information) requests to the Serbian Ministry of Interior (MOI) requesting locations of the cameras, including the analysis based on which these locations were determined, and details of the public procurement process and other relevant legal procedures.

After a little over a month, the Ministry replied that procurement documents were protected as “confidential.” The information on locations and analysis were not provided to us in any document or any medium. Here is a short article about it, including the attached FOI requests and related communications from SHARE Foundation.

Even though we didn’t get the exact information we requested, the denial in itself was valuable information and a key issue to present to the public. We got a strong reason to kick-off the public discussion and our research by writing about this denial and sending a press release to the few independent media outlets that instantly published the news. This starting process was important for us as it revealed that at times, the act of storytelling can be an act of investigation, or a way to kick-start an investigation.

By doing further research with publicly available information sources about this and similar projects worldwide, we found that for promotional reasons, Huawei actually published a case on their company website providing detailed information about the features and deployment of the smart surveillance system in Belgrade and its cooperation with the Serbian Ministry of Interior.

It was interesting (and somewhat shameful) that a private company served us so much more information than our own government. But this published case was also an opportunity for our background research. However, as soon as we analyzed newly obtained information and published it, the case study was removed from the Huawei website. We still managed to retrieve a snapshot of the webpage from Archive Today ( and the Wayback Machine (

Retrieving lost webpages and more

As this case demonstrates it, online archives are extremely useful resources for investigators, both for finding deleted webpages and for saving and preserving online information. Check the guide “Retrieving and Archiving Information from Websites available in this Kit, to learn more techniques, tips and tools that help you retrieve historical and ‘lost’ information from websites as well as archive and preserve your own copies of webpages for future reference. Image: Huawei case study on Serbia’s street surveillance needs. Source Wayback Machine capture

This shady and controversial sequence of actions by a public-private partnership confirmed that our government was hiding information from its citizens, but it also fueled the needed debate on this topic and helped us reach at least the initial bubble of concerned citizens through the media. We knew we had sparked discussion because people were commenting a lot and sharing the article in which we exposed this odd content removal.

At this point, everything we did was reactive, without much thought on strategy. What we felt, though, was that our government would just go on with the project. So, whether we somehow stoped it or not through a campaign, this fight would be long and hard and we could not go to battle alone.

Our next goal was to find a way to initiate direct communication with interested fellow-citizens, so we could continue these efforts together, and hopefully go beyond a one-off campaign and trigger some kind of an anti-mass surveillance movement that was in it for the long haul.

First contact: reaching out to the community

Belgrade is big enough to have a few individuals and small communities interested in topics such as privacy, security and surveillance. To be precise, this means two or three civil society organizations, one hackerspace, and several people occasionally tweeting about these issues. We initiated offline meetings to kick-start discussions and to come up with some ideas on how we should proceed. At first it was a small gathering in Belgrade Hacklab, then we got a workshop slot at a local hackathon-conference, which we used to promote the topic and reach out to some new people.

At this point, we had established communication with around 15 people directly, and reached out to 30 more, which were aware that we were being proactive about the issue. Most of them we already knew personally.

The initial campaign needs

The initial campaign “infrastructure” for the public campaign included:

  • creating a simple micro-site with a contact form,

  • choosing the website name and registering a domain,

  • running an initial social media push to build an organic (natural) community around the topic and campaign.

Now, we needed to reach out a bit further. To do this, we put up a simple one-page basic .html micro web-site containing only a few simple - yet strong - sentences and several links (as in the image below): Image source: SHARE Foundation / #hiljadekamera

Serbian to English translation of the text from the image above:

It said:

“As we speak, thousands of cameras with facial recognition technology are being deployed in Belgrade. Our police is hiding all information about the project. What are the locations, where is the data going, who has access and for what purpose?

This system will follow us when we commute to work, to see a friend, a doctor, grab a drink, go to a protest, light a cigarette, get out of the car, hug someone, talk, criticize.

Mass surveillance makes citizens obedient.

Do we want it?

If you want to stay informed or somehow help, let us know.

More about the topic: Blic / Peščanik / Foreign Policy / Telegram”

Being aware of the average human attention span nowadays, it was just enough content that everyone would read it, enough information to get to the point, enough links if someone wanted more, and most important, a link to the form where people could say what they think, offer support, help, and establish contact. Additionally, there was a link to a Telegram channel we created since privacy-aware people use it, and would feel familiar and identify with the anonymous source that put up the site.

When we finished the text for the micro-site, which involved making an ascii camera picture and putting it all together into a responsive html page that could be viewed well both from mobile and desktop devices, we needed an idea for an internet address (URL) that we could register. After mentioning “thousands of cameras” so many times those days, we realized that it was the perfect name for the URL, but also for a social media hashtag, a campaign and a movement. / #hiljadekamera (”thousands of cameras” in Serbian)

After someone from the team leaked the website we just wanted to preview on the internal Telegram group, it immediately got picked up by a well-known activist and went semi-viral. We expected this because we knew that there was already some interest in the topic. We got over 30 form submissions in no time. We didn’t get many Telegram channel subscribers, but it wasn’t so important since most of the people from the form left us their email, which was the best way to continue communication.

That first online wave went really well. Now that we had a name, a URL and micro-site, form and telegram channel, we could use that to go back to physical space and continue reaching out and building wider community.

Street marketing

Of course, stickers came first to our mind. But what kind of stickers?

Let’s not just print the URL or hashtag and mindlessly stick them everywhere. Especially in public places they would get removed by the city services. What we realized was that the police were required by law to mark areas that are under surveillance, but they were not doing it. This was our chance. We came up with the idea to disguise our stickers into official ones so we would:

    1. help citizens understand that they are under surveillance, and create a bit of “under surveillance” atmosphere downtown;

    1. apply our stickers to camera poles and not have them removed because they look official; and

    1. add a QR code which leads people to our website. Image source: SHARE Foundation / #hiljadekamera

We never knew how many people scanned the code, but the stickers were certainly popular giveaways at events, and filming the process of scanning the code with a phone and automatically opening our website turned out to be the favorite content for TV crews and journalists who were covering the story.

Bench under surveillance

Out of all the issues related to this project, the most abstract one to explain was how mass surveillance would affect citizen behavior and introduce self-censorship through “the chilling effect”, which refers to the inhibition of legal rights and is harmful for freedom of speech, freedom of assembly and democracy.

There is an annual open art festival in Belgrade, which means it’s pretty crowdsourced. Anyone can apply for a free spot and present their work, and almost anything can be considered art: a painting, a sculpture, acting and theater, workshops, playing music or mixing records.

Since our next idea was to explore and show how citizens are reacting to and feeling under surveillance, we chose to do a small installation and space intervention in order to embody the “chilling effect” among festival-goers by putting them under intrusive surveillance at times when they wanted to relax. Since there are usually no places where people can sit at crowded festivals, we got the most comfortable, best looking bench we could find and positioned it on a nice and cosy spot. Just over the bench, we deployed two dummy cameras with red blinking LEDs so they would appear real. Image source: SHARE Foundation / #hiljadekamera

The setting turned out nice. We just printed a QR code on a caption plate that was directing to a short questionnaire on how people felt while hanging out under surveillance. Questions were tailored to younger festival-going audiences and included something like: “Would you share secrets to a friend, would you dance, lie down, kiss or sext your partner under these cameras…?”

We did get some responses, but most of the interaction and discussion happened on the spot. We were surprised that youth could clearly express their feelings and were very aware of the mass surveillance intentions, mostly through popular culture - movies, books and video games - and that many of them were offering to join our mailing list and spread the word. However, most of them were still not aware of what it really meant for their privacy to have cameras creeping in every day and slowly covering streets of their city.

Engaging the community: a collaborative event

Our next steps were:

  • plan a datathon / sprint writing event,

  • choosing and tailoring information for the official website of the campaign.

By now it was almost a year since we were deep into research activities and discussions within the small existing community, with only occasional media outreach. At that point we finally felt ready to go further into campaigning mode and reach out to a wider audience.

For this purpose, we needed more than a obscure ASCII micro-site, stickers and a clear picture in our minds. We realized we needed a new push: a decent website with all the information we gathered so far, processed and systematized for a regular citizen, with several options for people to join the campaign.

It was important to include as much existing community members into this process for several reasons:

    1. creating more ambitious content and website requires different human resources and skills and we had no budget; and

    1. it was important to decentralize the campaign and make it as leader-less as possible to avoid a sense of ownership on our side, and to gain more engagement from all participants by making them feel equal partners in the project.

We made an early decision that the project / campaign would be independent and bias-free in terms financial support and geopolitics - we didn’t care if it was Chinese or European or American tech, so we didn’t want to receive any campaigning funds except from crowd-funding.

So, remember the online form we created and linked on the micro-site for citizens who were interested in staying in touch or helping out somehow? - We returned to it a week later only to find out it already had over 50 submissions of mostly messages of support, but also help offers from people with different skills such as visual artists, web designers, developers, researchers, journalists and lawyers.

This is where we came up with idea to invite them to a collaborative, sprint-like writing and design event where five groups would work together on different segments and content for a new website. Communication with participants was a challenge, but crucial for the success of the event. Initially, we invited at least 60 people. Half of them we already knew, some of them were friends of friends, and some of them we picked from the online form. As an event explainer, again we put up a single ASCII web page and sent it to potential participants as an invitation. It said:

“It has been a year since the Ministry of the Interior suddenly announced that the deployment of a mass video surveillance system with facial recognition technology was underway. The project and its implementation are still non-transparent. After many reactions, inquiries, several organised discussions and community work to reveal further information, it is time to gather around this problem and do as much as we can together.

The idea is:

  • to compile informative texts in several working groups,

  • to do a technical analysis of the system through patents and other publicly available information,

  • to enable open source camera mapping tools,

  • to create visual solutions for presenting data,

  • to create a plan of activities,

  • to tailor communications, and finally

  • to put it all together on the internet for fellow citizens to access.” Image source: SHARE Foundation / #hiljadekamera

After two weeks of planing and communicating, we had around 30 confirmations for the sprint event. It helped that we directly communicated with all of them, had a defined plan of what we wanted to achieve, and that the campaign was already gaining momentum. - Otherwise, it probably wouldn’t have been possible to get all those great people to spend a day volunteering on some abstract idea. Most of them knew what group they would be joining, but also that they could switch groups if they felt the need to.

Groups were named symbolically:

  1. .TXT - researchers and journalists collecting most the important information together and writing it into a human-readable language,

  2. .EXE group - mostly engineers figuring out how to best visualize and explain the architecture of smart surveillance systems,

  3. .PDF group - lawyers and advocates analyzing and summarizing all the legal issues about the project and strategizing at policy level,

  4. .MAP - developers working on an application for crowd-sourcing camera locations on a map,

  5. .HTML - web designers putting everything together into a new website.

For the event venue we booked a nice free community space gallery that had everything we needed to host this bunch for a whole day: tables, chairs, electricity, wi-fi and a toilet. We brought some croissants and coffee in the morning, had fruits all around the place and ordered pizza later in the day. We did not over-think or over-plan this action, we wanted it to be simple, comfortable and serve as a space for people to gather and feel safe talking about sensitive topics.

Prior to event, we sketched some ideas on what content the website should probably have and then briefly presented it to the team in the morning. Groups gathered around their tables and the sprint started. In eight active hours, we had all the textual content done, visual identity done, concept / sketch for a big surveillance system infographic, several iterations of mapping application, but most importantly, we had 30 people hanging out and working together towards one goal - to fight biometric mass surveillance. Most of us had just met, some of us knew each other but were collaborating for the first time, and many are still active on the project as of 2022.

Reaching beyond the bubble

In brief, this part was all about:

  • “the hunt” - crowd-sourcing a map of cameras,

  • “reclaim your face” - making a petition to ban biometric surveillance,

  • “ogether against thousands of cameras” - setting up a crowd-funding campaign.

Once we engaged the directly interested community, we wanted to go beyond and reach wider audiences to build more awareness and support.

After the sprint event, a month or two passed until we tightened it up and translated the content to English. The translation was important because we were obviously one of the first cases (guinea pigs) of the “western” world with an issue that other countries will also face sooner or later. So, our case is a valuable source of information and awareness on the topic and its developments - useful for researchers, journalists and activists around the world.

Once we released the website - - it was shared intensely on social media and we also got attention from traditional media, so TV and radio appearances followed. We knew the website and its content worked very well once we realized that these media appearances were going smoothly. - Newsrooms had everything in one place to do a story. All the leads and information editors needed, all the questions and answers for journalists were there, we just had to show up and talk it through. Finally, at the end of every appearance, when asked what citizens could do themselves, we were prepared for a concrete call to action: help us map the cameras around the city!

CTA “The Hunt”: crwodsourcing camera locations

The fact that the Ministry of Interior did not provide information about the project, nor the map and the exact number of cameras installed so far, was a good opportunity to “help” them do so, seek responsibility and make citizens aware of the mass surveillance issue.

After a clumsy attempt to create a crowdsourcing mapping application from scratch during the sprint event, we decided to use an existing open source application and just adapt it to our needs.

To avoid potential spam and problems with the relevance of the data we would receive from citizens, instead of inviting people to use the application itself, we asked them to send us photos of the surveillance cameras and their locations via Twitter with a hashtag, or directly via e-mail. This process required occasional administration of incoming content on our part, but it turned out not to be an overly demanding task, making our data management and verification much easier. Also, these kinds of mechanics turned out to be a good decision because the citizens were happy to publicly post photos on Twitter, which was a great campaign engagement, and promotion for the initiative’s Twitter account and its constant activity.

To promote this action, we again created a single webpage on the website of the initiative - - where we posted basic instructions, graphics with models of cameras that we “hunt”, and a temporary map. Camera maps and graphics with the sentence “how to recognize a camera that recognizes you?” went very well on social networks, but also on online portals and news agencies. The media liked the call to action and the fact that there were interesting graphic elements displayed. All this led to more media coverage of the campaign and our story as well as new guest appearances on TV programs. Image source: SHARE Foundation / #hiljadekamera

CTA “Reclaim Your Face”: petition to ban biometric mass surveillance

Many people are skeptical about petitions. There are few systems where they have legal weight and lead to direct changes. However, it turned out that their influence is more often indirect, and that petitions can actually contribute to exerting moral authority and spreading awareness about a certain problem. When colleagues from European Digital Rights (EDRi) suggested that we petition for a ban on biometric mass surveillance with several other organizations across Europe, we were immediately skeptical. After only a month, it turned out that this was one of the most important activities for us in the long run during the entire campaign.

Organizing a petition in the digital age can be easy. It is possible to create an account on one of the online petition platforms and then share that link across social networks. However, this simplicity has a downside:

  • First of all, there is great public saturation with the large amount of petitions on social networks, which makes it difficult for visibility and limits attention.

  • Second, you are inviting people to leave data on websites whose data and privacy policies you cannot control (many are selling email addresses, data leaks, cookies, trackers, etc.).

  • Third, you cannot control the interface and most of the content on these standard petition pages, thus functionality and public engagement suffer, because universal design that is made to suit everyone will never be as good as a unique design created for a special need.

Organizing a petition that way would be similar to promoting businesses via GeoCities (generic) websites, instead of having your own website.

Of course, creating your own website, a micro-site, a single-page site, and a unique petition platform requires resources. In addition to technical matters, good petitions require quality text, visual identity, design and production of different communication elements, promotion plan, etc. - What made the difference for us and our campaign was the importance of cooperation between like-minded organizations and membership in international organizations such as EDRi. Our petition was organized in several countries at the same time by local organizations, so most of the above assets were created collaboratively, and then each organization would only localize the materials.

Out of all the elements needed for a petition, the most challenging to produce and localize was probably the signature counter - a little piece of software where petition signatories need to enter their details. It had to be tailor-made, private by design (without trackers and with localized privacy policies), and then translated, color coded and finally embedded into each organization / country petition’s website. Once everything was in place, we simultaneously launched the petition.

Locally, we reached out to several supporters who could create some petition related content, which added up to the promotion. This is a work by a comic artist who covers daily news: Image source: Igor Lečić / Serbian to English translation of image text: “There is too many of them Sancho! / What should we do? / Let’s sign the petition!”

The petition feedback was incredible. Thousands of citizens signed up during the first several hours. Media outlets covered it and social media followed it. It was one of the trending topics for a day or two, which is great for awareness.

But why was this one of the most important activities of the campaign, as mentioned earlier?

While signing the petition, citizens would be able to opt-in for receiving news about the campaign. After only several days, we had almost 10,000 new subscribers on our mailing list, which from that point onward became our most important communication tool (up until that moment, we had only several hundred organic subscribers). We were not dependent on Facebook, Twitter and other gatekeepers to communicate anymore, at least not with all these people who opted-in (no dark-patterns!) and willingly gave us their email address to stay informed.

The petition also convinced us that we had increasing support. We expected to see 1000 signatures as a good result, but this was ten times bigger. It opened the door for the last activity of the year, ideal for the Christmas giving season - a crowd-funding campaign.

CTA ‘Together Against Thousands of Cameras”: the crowd-funding campaign

The campaign name and brand was already gaining recognition and a bit of coolness, so in collaboration with two local streetwear brands, we designed and printed some masks, beanies, bandanas and hoodies. This offering was yet another way for citizens to engage with the campaign, to give and show support. It led to yet another issue-awareness wave, and was also the first financial influx into the project - as we would’t seek or accept any funding from other organizations for independence reasons, as previously mentioned. This helped us continue all the efforts of battling mass surveillance. Image source: SHARE Foundation / #hiljadekamera

Crowdfunding, micro-donations, and donations towards civil society organisations (CSOs) in general are still not common in Serbia (as of 2021-2022) due to several factors such as weak individual economic power, late e-commerce adoption, strict financial policies and lack of interoperability with global payment services. Therefore, we set our crowdfundng goal at around 5000 Euro, which we almost reached in a day during a campaign pre-release for family and friends. This made us change the goal to 10,000 Euro, which we also reached far before the deadline. Once again we were surprised with the amount of support from our fellow citizens and were further encouraged to go on.

Multimedia and public relations

At this stage, we focused on:

  • multimedia production - short documentary, podcast and live-stream,

  • narrating media appearances - pandemic restrictions protest story and data cross-referencing examples.

Alongside all the activities described until now, an important part in communicating our work was the media content that we produced ourselves.

Different media formats will reach different audiences and will have varying success rates, so we tried with several outputs for which we had resources and people with skills to produce them. We created a podcast, a short video documentary and a live-stream event. Podcasting was an experimental form, part of an ongoing series following tech, society and cyberspace, so even if streamed only a couple thousand times (more via Telegram channels and RSS feed than YouTube), it was appreciated by niche communities, and thus important for the campaign brand awareness.

Our short video documentary was a huge success. We filmed cameras around the city, several actions we did including the sprint / datathon event, interviewed members of the initiative and the National Data Protection Commissioner. We were however ignored by the Ministry of Interior when we reached out with an interview request. Nice drone footage and dystopian music from local artists added a great deal to the professional direction, camera and editing done by our affiliates. The film’s outreach was great not only because it was shared on social media and watched online, but because it was aired on at least three cable TV stations multiple times in the span of a few months. Most of the cable TV stations will air content that they like and receive for free, especially if the format suits them well aesthetically, and they can fit it easily time-wise (ours was good for their needs as it was up to 15 minutes long.) A day after the documentary premiere we realized that once again we managed to break another bubble and reached even more fellow citizens.

Finally, in regular circumstances we would organise an event to promote our initiative’s work to date and to spark more debate. However, during pandemic lockdowns (this was 2020 - 2021), we were doomed with an online event. In the midst of the “Zoom fatigue”, we had to come up with something a bit more interesting in order to dive out of the over-saturated civil society webinars and online conference spaces.

Instead of inviting people to a video conference, we produced an online stream on YouTube, but made sure it was dynamic, visually appealing and welcoming. We had a host and five activists in the studio talking back to back on different segments of the surveillance issue and the campaign. We had four additional guests - a journalist, a security professional, a technology expert and a law expert - joining via video call, followed by a Questions & Answers session where viewers could address questions and comments. The production was done in our office, and included: Image source: SHARE Foundation / #hiljadekamera

In the end, the live stream was kinf of over-produced, since some of the watchers asked if this was a recording or a live-stream, but nevertheless it was an interesting way to talk about the matter and avoid another typical online event.

Basically, all of the above mentioned content and messages were spread organically, without any particular marketing and Public Relations strategy, and with zero funds invested on ads - our policy is to never pay Facebook, Google and the like for promotion.

Our philosophy was important, relevant topics and good content will find its way eventually.

Could the outreach with this content be better? Probably yes - but we were just a bunch of activists dealing impromptu with an emerging issue, on our own resources, not on anyone’s agenda.

It worked the same with public relations. We had many media appearances, but no-one is a professional PR / spokesperson in our team. Sometimes that’s stressful, having anxiety and tremor in front of cameras and journalists, in a live morning program or a podcast… But showing these vulnerabilities to a certain extent is also fine and can even be charming because no one expects activists to be professional spokespersons, or a grassroots initiative to contract someone like that. Mostly it’s just about good understanding of the issues and preparing several key messages aligned with actual events.

For example, massive protests in Belgrade in the pandemic year of 2020 were a strong example of how government is imposing fear through massive surveillance, and is using this system to suppress protest culture. At that time, government officials were constantly threatening in the media about using video footage and processing protesters, which made a lot of people change their minds about going out to the streets. This was a relevant occasion to further explain the “chilling effect” and it’s influence on society and protest turn-out.

Our key message when communicating during those days was that compared to protests of the last twenty years, these were the first occuring under thousands of cameras capable of recognizing our faces and converting them into data. This data is further easily stored in government data centers, ready to be cross-referenced with databases of people employed in the public sector, people who get social aid, or small business owners who might need additional tax officer inspection. Also, as our previous research proved, this kind of metadata is often retained forever, so even in five or ten years from now, someone could deny you a job, deny your child a scholarship or kindergarten admission, or make you a voter against your will because at some point in time you were out in the streets protesting government decisions you were opposed to.

Although it may sound dystopian, our research confirms that these surveillance systems combined with emerging and clumsy digitalization, data centralization, digital identity programs, smart cities and algorithmic governance lead to authoritarian societies, so these threats have to be communicated, and possible dark scenarios like these need to be narrated.

Once fellow citizens get to understand possible implications of these systems on security and privacy through examples, it is fair for them to choose whether they are for or against mass surveillance. But without information, different narratives and open discussion, it’s not the citizens who decide, but governments instead.

Finale: the first win

What happened so far:

  • public hearing where officials confirmed they will not use facial recognition, for now,

  • new internal affairs draft law legalising facial recognition proposed and withdrawn,

  • established community ready for the next battle.

The more our campaign grew, the less the Minister of Interior talked about the surveillance project. Actually, for more than a year they went totally silent on the matter. There was one leak of video footage from these surveillance cameras in the meantime that they instantly confirmed to us when we asked about it. For us, this meant more proof that the data was not secure, so we would also use the example whenever we communicated about the issue. On the other hand, we had no proof that authorities started using the facial recognition feature of the system, which aligned with recommendations of the National Data Protection Commissioner, but the Ministry never spoke about it.

Then, on 21 May 2021, during a public hearing about information security at a National Assembly, when asked about rising concerns of mass surveillance and facial recognition in Belgrade, a Ministy representative responded:

“We are aware that citizens are concerned about this issue, and we have witnessed over 200 articles during the last year about biometric surveillance in Belgrade. I can assure you that, even if some people from the Ministry of Interior think that it is possible by current law, until there is a public debate, until there is consent, and the Assembly vote, we will not employ software for facial recognition.”

Three months later, the Minister of Interior announced the new Draft Law on Internal Affairs, one that, among other things, contained provisions for legalising a massive biometric surveillance system. It was announced under the radar on the Ministry’s website, without the media being informed, with less than three weeks’ deadline to submit comments. We were lucky to even find out about it a week before the deadline, but we were prepared.

We joined a public discussion event (the first one ever organized by the government on this topic), sent our comments to the Ministry, and alerted international and community media. Even though heated public debate followed for several days, we were sure that it wouldn’t affect the government’s intention, and that accepting comments from civil society was just a formality before passing the law.

Four days after the deadline and with strong public opposition, an unexpected press release was issued by the Minister himself:

“Considering all the facts, President Vucic asked me to withdraw the draft law on internal affairs from the procedure. I will do it. That is my defeat, and I will bear consequences, but I also want to let the public know that I will continue to fight for a safe Serbia despite the (foreign) agents and their financiers. The draft law on internal affairs has been withdrawn; you will have to look for some other reason for the blood on the streets of Belgrade.”

It seemed incredible, but after that quote at the beginning of this article where the Minister himself was assuring the public that nothing will be able to deter new smart cameras that were already being deployed in the streets of Belgrade, two and a half years later, after all the efforts we put to prove it’s not legal and we don’t want them, the law to legalise them was proposed and then urgently withdrawn.

But even more importantly, all these efforts, done together by the activists and joined by many fellow citizens, made a strong community. A community with its own narrative, one that protects the integrity of every citizen, a community ready for the next challenge.

Published November 2022


Technopolice resources - tools and methods from French initiatives providing ways to leak files for people who work in companies involved in Biometric Surveillance and administrations (content in French).

Technopolice Campaigning Action Guide (content in French).



ASCII art - a graphic design technique and aesthetics for presentation of pictures pieced together from basic computer printed characters. (Wikipedia.


Biometric surveillance - a technology that measures and analyzes human physical and/or behavioral characteristics for authentication, identification, or screening purposes. (Wikipedia)


Chilling effect - an effect that reduces, suppresses, discourages, delays, or otherwise retards reporting concerns of any kind. (Wikipedia)


Counter-narratives - stories impacting on social settings that stand opposed to (perceived) dominant and powerful master-narratives. The concept of counter-narratives covers resistance and opposition as told and framed by individuals and social groups. (Source: Routledge Handbook of Counter-Narratives 2021)


False narrative - a story that is perceived as being true but has little basis in reality.


Street marketing - a form of guerrilla marketing that uses nontraditional or unconventional methods to promote a product or service. (Wikipedia)


Panopticon - the idea of total surveillance centralization is older than sophisticated computer systems and originates from the new prison design proposal dating back to the end of the 18th century. According to the same design, the daily life of prisoners is monitored from the central tower, from which someone can but does not have to be watching them. The very idea that they can be spotted at any point, so that prisoners always behave as if they are being watched. (Source: hiljadekamera).


Political narrative - a term used in the humanities and political sciences to describe the way in which storytelling can shape fact and impact on understandings of reality. However, political narrative is not only a theoretical concept, it is also a tool employed by political figures in order to construct the perspectives of people within their environment and alter relationships between social groups and individuals. (Wikipedia)


**Responsive html - Responsive web design (RWD) is an approach to web design that makes web pages render well on a variety of devices and window or screen sizes from minimum to maximum display size. (Wikipedia).


Viral (in internet terms) - The term viral pertains to a video, image, or written content spreading to numerous online users within a short time period. (Wikipedia)