Abstract
This paper examines how visual communication is becoming a key aspect of migration governance, particularly in shaping accountability within border policing. Focusing on Frontex, the European Union’s Border and Coast Guard agency, this study analyses how the agency uses images on X (formerly Twitter) to define its ‘account ability,’ signalling its competence and responsibilities. Through a visual intertextual approach, which considers both images and accompanying text, the paper analyses a collection of 393 images from 323 posts between 2015 and 2021 of Frontex’s work at sea. Using PixPlot to create clusters similar images, followed by further selection and creation of montage of images, the paper reveals how Frontex uses visuals to establish the boundaries of its accountability in border policing. The paper presents two case studies, one of its images in relation to search and rescue, and the other of its crime fighting prowess, to examine how Frontex’s images on X work as a form of visual governance in migration.
Introduction
The European Border and Coast Guard Agency (Frontex) has a considerable presence on social media. The agency posts images and videos of its work across Facebook, YouTube, Instagram, and LinkedIn. Frontex is particularly active on X (formerly Twitter), where it has the most followers (around 66,000) and has shared thousands of posts since opening its account in 2015. Its presence on social media is a key mechanism for the agency to demonstrate accountability and thereby bolster its legitimacy. This paper examines how Frontex utilises X to disseminate images that demonstrate and delimit its accountability for its work at sea.
Frontex emerged as a response to help coordinate border policing across the EU. All EU member states are obliged to join the Schengen area, except for Ireland, which has an opt-out approved by the European Council. The Schengen area, in principle, allows for the free movement of people between 29 countries. The removal of internal border controls among these states necessitates coordination at a secure and shared external border (Ceccorulli, 2019: 304). The EU and its member states share the responsibility for managing border security. The EU has the competency to develop uniform policies in border management. Each member state ultimately remains responsible for how border policing operates on the ground. They may legislate on matters relating to borders where the EU has not already legislated (Mungianu, 2016: 20). Frontex, an EU agency established in 2004, helps coordinate and support the border policing efforts of EU member states.
Frontex supports EU member states and the four Schengen-associated countries, which are European countries that are not EU members but are part of the Schengen area, in all facets of border policing. The agency provides operational, financial, surveillance, and coordination support to the border policing efforts of EU member states and Schengen-associated countries. Frontex also conducts joint operations with third countries, particularly in the Western Balkans (e.g., Albania, Montenegro, Serbia). The agency has its own assets to conduct surveillance of the EU’s shared external border and beyond (known as the pre-frontier area). Frontex uses surveillance from satellites, radars, and cameras on planes to beam live footage back to the agency’s headquarters in Warsaw. Frontex also monitors trends in migration, collating and sharing trends in border crossings. Lastly, Frontex provides various training sessions for border guards, including those on detecting fraudulent documents and human rights training. Frontex is the first of its kind: a transnational border policing agency with its own uniformed officers who have executive powers to use force along the EU’s external border.
Accompanying this broad mandate is significant financial backing from the EU. Frontex is the most well-funded EU agency. In 2023, it received a budget of €845 million (Frontex, 2023). Historically, the largest share of Frontex’s budgets has been allocated to its maritime operations (Frontex, 2012b: 26; House of Lords, 2008: 32). The high costs of leasing and operating vessels or aircrafts for its maritime surveillance make Frontex’s joint operations at sea particularly expensive (Frontex, 2012a: 26). In 2015, the year of the so-called migration crisis, 62% of Frontex’s €113.5 million operational budget was spent on its joint operations at sea (Frontex, 2016: 40). Frontex’s work at sea accounts for most of the agency’s budget, and yet, very little is known about how these operations are implemented.
Frontex either denies the public release of documents or heavily redacts operational plans released in response to freedom of information requests. The agency denies and redacts this information on the basis that releasing such details would compromise the security and efficacy of ongoing operations (Gkliati and Kilpatrick, 2022: 65). Frontex is the most well-funded EU agency, with the majority of its funds allocated to maritime operations.
Hence, there is a clear impetus for transparency and accountability regarding how Frontex operates at sea and what its maritime operations aim to achieve. Reports from civil society organisations have drawn attention to the human rights risks linked to Frontex’s maritime activities, reinforcing calls for greater accountability in its policing of the EU’s sea borders (Amnesty International, 2023; Médecins Sans Frontières, 2024). Frontex must account to EU citizens for how it spends public funds and respond to public scrutiny of its actions. All while maintaining control over the agency’s operational details.
On X (formerly Twitter), Frontex carefully crafts its accountability for its work at sea. Numerous images of Frontex officers and assets at sea appear when scrolling through Frontex’s X page. These images peel back the curtain on a facet of Frontex’s work that has limited visibility, and yet a high degree of controversy. The images Frontex posts on X offer a type of visual accounting for its activities at sea (Davison, 2014). One that allows Frontex to curate its accountability for its role in sea operations, its most well-financed and controversial component of its work. Frontex uses X as a form of visual governance to communicate and legitimise its policing practices (Massari, in this special issue).
In this paper, I use visual methods to unpack the discourse surrounding accountability for border policing. The paper builds on the existing literature on migration and visual methods, policing and social media use, and Frontex and accountability. A growing body of literature examines how the press portrays images of people migrating to Europe (Chouliaraki and Stolic, 2017; Hansen et al., 2021), as well as how individuals moving across borders perceive themselves (Risam, 2018). Policing scholars have highlighted the role of social media, particularly X, as a tool for reporting and advancing the aims of border policing and policing agencies more broadly (Kane, 2024; Walsh, 2019; Walsh and O’Connor, 2019). Finally, doctrinal legal scholars have highlighted the challenges in holding Frontex accountable when a possible human rights violation occurs during a joint operation (Fink, 2018; Gkliati and Rosenfeldt, 2018; Karamanidou and Kasparek, 2020; Pollak and Slominski, 2009). Nonetheless, there is scant research on the intersection of these topics. Specifically, how images posted on X by a border policing agency serve as a mechanism for accountability and how this reasserts dominant discourses surrounding migration to the EU.
In this paper, I analyse a collection of images Frontex has posted of its work at sea. I present two case studies, one on search and rescue and another on crime fighting, that demonstrate how Frontex uses images to represent itself as accountable and effective. Before presenting the case studies, I clarify my use of concepts, the existing literature on Frontex and its use of images on X, and the visual methods I used to gather and analyse the collection of images. Following the case studies is a brief discussion on the inherent challenges Frontex faces as a transnational policing agency, and how this influences its audiences and image work.
Frontex’s image work & accountability
Visual methods have been a powerful tool for holding Frontex accountable for its human rights obligations. An October 2020 investigation from journalists at Lighthouse Reports and Bellingcat published videos and images of Frontex-financed boats participating in an illegal pushback (Giorgos Christides et al., 2020). This investigation ignited reviews from Frontex’s Management Board and led to the development of the European Parliament’s standing Frontex Scrutiny Working Group. While journalists and civil society organisations have used cameras to improve the agency’s accountability for human rights, Frontex has also been using cameras to define what it ought to be accountable for.
Richard Ericson (1995) referred to this type of work done by policing agencies as a way of strategically negotiating ‘account ability’. For Ericson (1995:137), account ability is ‘the capacity to provide a record of activities that credibly explains them so that they appear to satisfy the rights and obligations of accountability.’ Crucial to an organisation’s ability to account is the medium through which it conveys its ability to fulfil its responsibilities. Different mediums are used to present various stories of answerability. Narratives about policing in newspapers differ from those offered on television.
Nevertheless, visual elements in both formats share a common bond. In Ericson (1995)’s view, the police use the media to relay stories that account for their job as effective, allowing the public to identify with the work and mandate of the officers. By providing an accounting of their actions, policing agencies shape the very parameters of what they should be accountable for.
Such public relations activities have been referred to as ‘image work’ in the policing literature (Mawby, 2013; Mawby and Reiner, 1998). The ubiquitous use of cell phones and access to social media have led to a new visibility for policing, and a myriad of mediums for the police to engage in image work (Goldsmith, 2010; Haggerty and Sandhu, 2014; Sandhu and Haggerty, 2015). What were previously backstage policing activities, out of sight and scrutiny from the public, now have the possibility of becoming front-stage and visible to the public (Goldsmith, 2010). With police work becoming increasingly visible, Frontex has seized the opportunity to enhance its public image on social media platforms.
To date, three studies have analysed the visuals produced by Frontex. A recent study by Alice Massari (2024) examined 117 images from Frontex’s X account to compare them with those of three other migration institutions operating in the EU and Canada. Massari found that Frontex’s images centre on white men in their 30s and 50s, and place border guards or senior management as central characters (Massari, 2024: 12). Massari (2024: 14) also found that Frontex decentres the border, treating it as a backdrop to the activities of its staff. The second study from Katja Franko (2021) included a textual and visual analysis of 525 images and 64 videos posted on X by Frontex and Médecins Sans Frontières. Franko (2021:393) offers a counterargument to the notion that Frontex’s activities represent a merging of humanitarian and policing practices, as there are stark differences between Frontex and Médecins Sans Frontières in the discourse surrounding search-and-rescue operations in the Mediterranean (Pallister-Wilkins, 2015). Lastly, a doctoral dissertation inductively examined a sample of 554 images from Frontex’s annual, general, and risk reports between 2010 and 2016 (Silberstein, 2020: 42–43). The images reproduced subject positions, such as ‘irregular migrants’ or vulnerable children, and normalised Frontex as a necessary response to security threats (Silberstein, 2020: 199).
The visual material used in all three studies offered new perspectives on Frontex’s communications and the agency’s role in shaping discourses on migration. The images revealed the performative elements of securitising the EU border, and who is and is not protected. As I demonstrate in this paper, the images Frontex uses are also strategic inputs and outputs of visual governance (Massari, in this special issue). Frontex posts images on X to shape public perceptions of its own efficacy at sea, and curate its account ability.
Methods
In an age where more new photographs are uploaded daily than could ever be analysed in a lifetime, it is helpful to examine online images in groupings and collections, rather than conduct close readings of online photos in isolation from a group (Colombo et al., 2023: 1957). Social media images are notable for their multiplicity, circulation, modification, ‘networkedness,’ and platform specificity (Colombo et al., 2023: 1957). For this paper, I have grouped Frontex’s images of its work at sea to understand its visual grammar and how it has evolved. The data collection and analysis methods I describe below are my adaptation from the literature on visual methods, combined with digital research techniques I learned during a collaborative project at the University of Amsterdam’s 2023 Digital Methods Summer School (Niederer and Colombo, 2024; Rogers, 2013).
Data collection
I began data collection by deconstructing each post Frontex uploaded on X between 2015 and 2021. I assigned a unique ID to each image that connected to the text of the X post. In total, I collected 2171 individual posts on X and 1912 images.
Using PixPlot, an open-source and web-based software, I grouped the 1912 images by their visual similarity. PixPlot uses a convolutional neural network and Uniform Manifold Approximation and Projection (UMAP) to cluster images that appear similar to one another in a web browser (Andy Coenen and Adam Pearce, n.d.; Yale DHLab - PixPlot, n.d.). PixPlot creates groupings based on visual similarities, efficiently producing extensive collections of images. While UMAP introduces an element of computational error, it reduces human error and subjectivity when selecting similar images within large data sets.
Figure 1 displays the output I used from PixPlot to select a cluster. The output shows a global view of all the pictures Frontex has posted on its X account between 2015 and 2021. Using the lasso feature, I circled the clusters. I conducted a closer review of the clusters, verifying their coherence, and removing images that were incorrectly joined due to computational error. Ultimately, I selected cluster 1, comprising a verified sample of 398 images of Frontex officers or assets at work at sea.
1
A cluster map using PixPlot of all the images Frontex has posted on X (formerly Twitter) between 2015 and 2021. I hand-drew the lines around the clusters to indicate how cluster 1 was chosen and analysed for this study.
Clusters 6 and 7 in Figure 1 contained maps and infographics that also occasionally cited Frontex’s work at sea. There may also be posts on X about Frontex’s work at sea that do not contain images of boats and officers near large bodies of water, as I have filtered for this study. I chose only to focus on how Frontex portrays itself at sea borders. Kress and Leeuwen (2021: 118) note that maps and diagrams deemphasise gaze, and ‘value objective, dispassionate knowledge, ostensibly free of emotive involvement and subjectivity.’ Although Frontex’s maps and diagrams add to the agency’s discourse on migration by irregular routes as akin to invasion (Van Houtum and Bueno Lacy, 2020), I am focused on how Frontex portrays itself at sea. What Frontex centres and omits in its images of itself at sea, and how this reflects the agency’s idea of what it should be accountable for in border policing at sea.
Data analysis
I adopted a visual intertextual approach to analysing the data. This technique accounts for the potential bias of ocularcentrism, which is the privileging of sight and images over other senses and forms of expression, such as hearing and narrative forms (Jay, 1988). Martin Jay (1988: 323) wrote extensively on ocularcentrism, and argued that where visuals and texts are together, ‘…it is the complex interaction of the two that one of the great motors of human culture can be discerned.’ The text and visual expressions work together to allude to other social, political, and cultural events or processes. An intertextual approach to examining posts on X is imperative, considering that when scrolling on X, the text precedes any image. The text sits above the image(s) on an X post. While an individual may glance at the image or thumbnail before the text, it is the text itself of the post which situates the role and function of the image.
To implement a visual intertextual approach, I iteratively coded both the images and the corresponding text. I bounced between looking at the image and the corresponding text of the tweet. I coded each image and its corresponding text together to themes that illuminated both the content of the text and the elements within the image. I began coding by looking at the pictures and taking note of their denoted meaning, which is the literal, objective representation of what is in the image (Barthes, 2007: 192). Once the essential content of the images and texts was coded, I created smaller collections of images for the two most prominent themes: search and rescue and crime fighting.
I refer to these smaller collections as montages, with each montage containing 5 to 10 images that represent Frontex’s depiction of its work in search and rescue or crime fighting. Some of the images in the montages also induce a sense of punctum, pulling me in emotionally, or contain elements of action evoking a multisensory response about policing at sea (Barthes, 2010; Pink, 2011). Other images were selected for a montage because they clearly expressed one of the sub-themes. Once the montages were created, I also sought the connotative meaning, which refers to any cultural and interpretive associations that the image evokes (Barthes, 2007: 192).
The montages are made to be read like filmstrips. In each row, beginning from the top left and moving to the right. The images are organised chronologically in ascending order, with older posts on the left and newer ones to the right. The montages serve two purposes: (1) to present a visual summary of the findings; and (2) as an analytical tool to reflect on how the selected images work together as a collection, making them more than the sum of their parts.
When reflecting on each montage, I also asked myself what these photos desired from the beholder (Mitchell, 2005: 33). I saw the montages as a set of repeated demands by Frontex to an audience. Rather than looking at the montages and asking myself, ‘what do these pictures mean?’ I looked over the montages and asked, ‘what do the pictures want from us?’ (Mitchell, 2005: 49). Although Mitchell (2005) does not offer a methodological roadmap for answering this question, he provides additional probes that may guide the reflection, such as: examining the potential target audience; noting any missing elements in the picture; and noting the possible things excluded by the boundaries of the image (Mitchell, 2005: 49–50). I reflected on these questions as I looked over the montages in my analysis.
Posting effective & affective images of border policing
Key Themes in Frontex’s Posts on X when Sharing a Picture of its Policing Activity at Sea (2015-2021).
Between 2015 and 2021, the majority of posts that Frontex shared on X regarding its work at sea focused on search and rescue operations. Among the 323 Frontex posts containing at least one image of staff or assets at sea, 40% concerned search and rescue operations. The majority of these posts on search and rescue at sea (60%) were uploaded between 2015 and 2016, coinciding the so-called migration crisis. Mentions of search and rescue at sea dropped precipitously after 2017. From 2018 onwards, whenever Frontex posted an image of its work at sea, it was most often accompanied by text explaining its role in combating crime. The case studies below demonstrate how Frontex utilises images to prove its effectiveness in search and rescue and crime fighting, as well as to elicit emotional responses that garner sympathy for its work.
Search & rescue at sea
Border policing at a distance: Overcrowded & unseaworthy vessels
When posting on X about search and rescue at sea, Frontex frequently uses images from its maritime aerial surveillance program. This program collects data from cameras and sensors placed on aircraft and drones, which fly past the maritime boundaries of EU member states. This surveillance equipment is aimed at detecting suspicious and unseaworthy vessels. The airplanes and drones stream the video from their cameras, as well as send images and sensor data back to Frontex’s headquarters in Warsaw. All of this data is then uploaded to EUROSUR, providing a live 24/7 situational picture of the EU’s external borders. The images Frontex shared on its X account, as shown in Figure 2, have all the hallmarks of its maritime aerial surveillance program. A montage of images included in Frontex’s posts on X (formerly Twitter) search and rescue at sea with overhead and drone shots of overcrowded vessels (2015-2021). This montage was created using ImageJ correcting for varying image sizes.
2
What stands out amongst the images and text that Frontex shares on search and rescue is that they emphasise how the vessels used are unseaworthy, requiring interdiction at sea. Image A1 in Figure 2 was accompanied by the following text: ‘Croatian vessel Adrija Mohorovičić rescues 236 migrants off a rusty fishing boat as part of Operation #Triton’ (Frontex [@Frontex], 2015a, emphasis added). Similarly, image A2 in Figure 2 was posted by Frontex alongside the following text: ‘Vessel, aircraft from Frontex’s Op. Indalo help in rescue of 15 migrants on a sinking rubber dinghy south of Spain,’ (Frontex [@Frontex], 2015e, emphasis added). Together, the text and images reinforce how the vessels are fraught with risk and require an intervention from Frontex. In uploading these images, Frontex justifies its surveillance activities as necessary to detect dangerous situations at sea and alert authorities to intervene against sinking rubber dinghies and rusty fishing boats.
Frontex is carefully defining and pruning its role in search and rescue through the images it shares on X. The posts focus on Frontex’s capabilities to detect and alert search and rescue authorities. However, the posts make little mention of Frontex’s obligations to protect the rights of the individuals on the unseaworthy vessels. The right to leave a country, seek asylum, to life and liberty, and against non-refoulement, are protected under human rights instruments that Frontex is obligated to comply with in its operations (European Parliament and Council, 2019; Guild, 2013). Rarely does Frontex make note of these obligations on its X account.
Where individuals are disembarked is a key indicator of whether human rights are protected and whether individuals can exercise their rights. Where a vessel is in distress, the rescuing party must ensure that the individuals rescued are taken to a place of safety (Trevisanut, 2014: 6). A place of safety is one where ‘the rescued persons’ safety of life is no longer threatened; basic human needs (such as food, shelter and medical needs) can be met; and transportation arrangements can be made for the rescued persons’ next or final destination’ (International Maritime Organization et al., 2015: 13).
The descriptions Frontex provides of its search and rescue activities often exclude where the individuals were disembarked. Between 2015 and 2021, Frontex shared 99 posts that reported on specific operations or recent rescue events involving Frontex. Nearly one-third of these posts (n = 31) lacked details of the maritime boundaries within which the rescue occurred, or provided no clue as to whether the individuals were disembarked to a place of safety. The text accompanying image B3 in Figure 2 demonstrates how Frontex uploads images about search and rescue, but omits details of where the individuals were disembarked, as the post stated, ‘Frontex Executive Director Fabrice Leggeri: “The tireless Frontex surveillance team works day and night. They again played an important role in the rescue of more than a 100 people and helped to avoid more casualties,”’ (Frontex [@Frontex], 2020a).
Additionally, seven posts mentioned that the search and rescue operation occurred in Libyan waters or was carried out by the Libyan Coast Guard. Civil society organisations and a UN fact-finding mission have long warned that Libya is an unsafe place to disembark (Amnesty International, 2021; Human Rights Watch, 2017; United Nations Human Rights Council, 2021). By omitting—and in the case of Libya, disregarding whether the individuals Frontex helped rescue were disembarked to a place of safety—Frontex uses X to narrow its responsibilities in search and rescue. Frontex delimits its role to detecting unseaworthy vessels and alerting nearby authorities to rescue or intercept the vessel. In doing so, Frontex is distancing itself from the responsibility to ensure that irregular migrants are taken to a place of safety.
Border policing close up: Action & emotion
The previous section showcased images taken from above and at a distance from individuals who travel irregularly. During 2015-16, the time of the so-called migration crisis, Lampedusa, Italy, became the theatrical setting of the unfolding humanitarian crisis (De Lauri and Brkovic, 2024: 8). During this time, Frontex also presented a dramatised image of its role in border policing on X. In its images, Frontex brought the viewer closer to border policing at sea. Examples of these images are included in row A of Figure 3. These images put the viewer into the position of an ‘invisible onlooker’ who can look at the journeys of irregular travellers and the work of Frontex officers, as though they were on display (Kress and Leeuwen, 2021: 117). A montage of images included in Frontex’s posts on X (formerly Twitter) on search and rescue at sea with Frontex staff performing dramatic search and rescue and in action (2015–2021). This montage was created using ImageJ correcting for varying image sizes.
The close-up images in Figure 3 not only present a firsthand account of how Frontex performs its border policing at sea, but also conjure up particular emotions towards the subjects. Figure 3 showcases images that portray Frontex’s search and rescue work as heroic and thrilling—sentiments the agency deliberately promotes on its social media platforms (Kilpatrick, 2021). Row A in Figure 3 shows images of toddlers being held by Frontex staff after they were rescued from unseaworthy vessels. Between 2015 and 16, a woman or child being rescued from a boat appeared in nearly a quarter of all of Frontex’s posts related to its work at sea. These images present irregular travellers as vulnerable and in need of rescuing, while simultaneously showcasing Frontex as heroic in responding to an emergency. The choice to make Image A1—depicting a child being rescued—the agency’s first post on X highlights Frontex’s intent to project a valiant public image from the outset. (Frontex [@Frontex], 2015b).
Where Frontex posted an image of a child or woman, it was often accompanied by text indicating the number of individuals rescued and a hashtag referencing the name of Frontex’s operation. For instance, Image A2 in Figure 3 was depicts a Frontex officer rescuing an infant. The image was reposted by Frontex alongside the text, ‘Norwegian Sim Pilot brought 407 rescued migrants to Catania. #Triton’(Frontex [@Frontex], 2015c, emphasis added). While the text often did not specify if women and children were rescued, the image used ensured that the visual that came to mind was that of vulnerable persons, relaying the efficacy and indispensable quality of Frontex’s work at sea.
Pictures of toddlers at sea do not automatically frame irregular travellers as vulnerable and deserving of protection. Images like the ones in row A of Figure 3 may contribute to a discourse that categorises parents who take their children with them on irregular routes as ‘fundamentally flawed in their capacity as parents’ (Stumpf, 2020: 1072). Such images can lay the seeds for policies that prevent departure, thereby protecting the children from the seemingly flawed decision-making of their parents. This discourse on flawed parenting can reinforce Frontex’s priorities of detecting and preventing the departure of irregular travellers by sea. The images, like those in row A of Figure 3, omit any discussion of how the EU’s migration policies restrict safe pathways for migration, leaving these perilous journeys as the only option for individuals escaping persecution or economic hardship in their home country.
The images in row B of Figure 3 also relate to search and rescue at sea, except they centre Frontex staff and assets in the frame. The images dramatize search and rescue at sea, presenting a thrilling and dangerous environment where Frontex heroically intervenes to save lives. The emphasis is on the action within the frame, rather than on who is necessarily in the frame. In B1 of Figure 3, a coast guard throws a ring buoy. In B2 of Figure 3, some border guards save a ‘Pakistani man having heart attack’ on board a ship (Frontex [@Frontex], 2015d). The text accompanying B3 in Figure 4 further emphasises Frontex’s work as heroic and exciting, stating that: Regular exercises help our officers to polish their skills and work together to assist people in distress. In a rescue operation, every second counts. This is why the exercise included emergency life-saving techniques and safe disembarkation of people once they reach land (Frontex [@Frontex], 2020b). A montage of images included in Frontex’s posts on X (formerly Twitter) on detecting and seizing drug across sea (2015-2021). This montage was created using ImageJ correcting for varying image sizes.
Together, the images and texts in row B of Figure 3 attempt to depict Frontex’s work as thrilling, high-speed, and dangerous. Such representations of policing have been referred to as ‘crediting images,’ whereby by the police are depicted as heroic in fighting crime or performing rescues, (Sandhu and Haggerty, 2015: 8). Crediting images are necessary to combat pictures which may be critical of the police, such as video or photographs showing an unreasonable use of force (id. ibid).
These types of crediting images also combat critical discourse about a policing agency by positioning the police as effective. The pictures in row B of Figure 3 are in contrast to research that shows the day to day practice of border policing—like local policing— can be filled with routine practices and boredom (Fabritius, 2024; Konopinski, 2014; Phillips, 2016; Rowe and Rowe, 2021; Sausdal, 2021). The images omit the boredom associated with search and rescue, such as waiting for a port to disembark, or the time spent waiting to detect a vessel in distress. These images are likely also used as a recruiting tool to attract new hires, presenting Frontex as an exciting agency to work for. In all, the photos in row B of Figure 3 attempt to give the appearance that Frontex is effective. The images in Row B of Figure 3 serve as a visual rebuttal to discrediting images or data that might question Frontex’s operational success.
Finally, the emotional importance of images in building and shaping discourses that support how Frontex operates and what it ought to be accountable for is exemplified in the text that accompanies images A5 and B5 of Figure 3. These two images were posted alongside a peculiar text that signalled the extent to which Frontex values images that evoke emotions and heroism. Frontex posted one photo nearly every week from late August to December of 2019 with the caption: ‘
Photo of the week 
.’ Images A5 and B5 in Figure 3 were part of Frontex’s photo of the week collection.
The image of the toddler being rescued and held, along with the action shot of a dangerously tilted Frontex-funded vessel, resonate emotionally, eliciting sentiments of heroism and effectiveness. The photo of the week collection are visual markers that Frontex is competent in fulfilling its search and rescue responsibilities. The posts omit key details, such as whether the agency disembarks at a place of safety or if its patrols at sea prevent access to safer routes. The photos of the week combat this criticism, providing crediting images of Frontex as capable, heroic, and impactful.
Crime fighting at sea
In contrast to the previous section, which depicted the sea as dangerous for irregular travellers, the posts I present in this section represent the sea as a site which is hazardous for Europeans. The sea in these pictures is characterised as a space that is rife with human smugglers and drug traffickers.
Between 2015 and 2021, Frontex shared 130 posts discussing search and rescue at sea, compared to 60 on crime fighting at sea. However, from 2019 onwards, Frontex’s posts of its officers and assets working at sea centred more on crime fighting at sea than search and rescue. Between 2019 and 2021, Frontex posted 24 images of itself detecting and intercepting human smugglers, compared to 12 images of itself performing search and rescue at sea. From 2019 to 2021, Frontex presented itself more visually as an effective crime-fighting agency than as one that conducts search and rescue operations.
Drug smuggling
The images in row A of Figure 4 emulate the image work that local police do in accounting for their effectiveness. The images in Row A of Figure 4 depict vessels intercepted with contraband on board. The pictures in row A reproduce elements that are familiar in policing television and film: action, bravery, and a clear separation between heroes and villains (Dowler, 2016: 23–24). As discussed in the previous section, Frontex’s search and rescue posts can evoke sympathy for both the agency (as hero) and irregular travellers (as victims). In contrast, the images and accompanying text in crime-fighting posts are intended to generate sympathy exclusively for Frontex. This framing positions drug and human smugglers as villains and casts Frontex in the role of hero.
The images in row A emphasize the connection between movement, action, and heroism (Pink, 2011). Image A3 in Figure 4 includes a caption describing a speedboat corralled by border authorities, potentially avoiding a chase at sea (Frontex [@Frontex], 2017b). Similarly, the text accompanying image A5 in Figure 4 reported that “Italian authorities chased down both boats…” (Frontex [@Frontex], 2018, emphasis added). The images in row A, depicting high-speed chases at sea to track down drug smugglers, convey excitement and effectiveness in combating crime at sea.
Frontex also depicts itself as an effective crime-fighting agency by uploading police trophy shots, as shown in row B of Figure 4. Police trophy shots are photos commemorating an arrest or drug bust (Linnemann, 2017: 3). Katja Franko (2021: 389) has previously highlighted Frontex’s use of trophy shots on X to ‘communicate victory and its hunting prowess.’ Images of intercepted contraband and arrests function as police trophy shots, creating high-speed chase narratives that bring viewers into close proximity with the action. Frontex situates its work at sea as a dangerous space where criminal enterprises would run rampant without the agency’s capabilities to detect and intercept drug traffickers. Frontex shares these trophy shots as material evidence that the agency is capable and effective in combating drug smuggling.
Human smuggling
Unlike the simple binary of heroes and villains that Frontex presents in its posts on drug smuggling, Frontex’s posts on human smuggling try to reflect multiple aims. In several instances, Frontex’s reporting on human smuggling at sea was conflated with search and rescue operations. The images in row A of Figure 5 are indistinguishable from the photos in row A of Figure 2 presented earlier in this paper. Row A of Figure 2 drew attention to the need for the detection and interdiction of unseaworthy vessels. Conversely, in row A of Figure 5, Frontex utilises a similar aesthetics when relaying statistics and instances of human smuggling. Image B1 in Figure 5 is the first post from Frontex where it mentioned human smuggling as part of its maritime operations. The photo is taken at eye level, in close proximity to the arrest. However, from 2017 onward, Frontex began using the same bird’s-eye view perspective it had employed in its search and rescue imagery to depict arrests related to human smuggling. By applying the same visual aesthetic for posts related to search and rescue to fighting human smuggling, Frontex is using X to frame all vessels as potential threats. Applying the same aesthetic and angle conflates individuals fleeing persecution or economic hardship with criminality, associating them with human smugglers.
In four posts, Frontex combined references to search and rescue with statements about arresting or denouncing smugglers. In doing so, Frontex subtly reframes humanitarian efforts as operations tainted by criminality. One post for example stated, ‘Frontex Director on Olympic Commander: Norwegians show what Frontex is about. Their crews rescued 35,000, helped arrest 200 smugglers’ #Triton’ (Frontex [@Frontex], 2017a, emphasis added).
The same aerial perspective once used to evoke vulnerability is now used to criminalise. Frontex’s framing omits key details of how the agency determined who the smugglers were on board. The omission contributes to giving smugglers an ‘evanescent character,’ making them challenging to find and fight (Achilli and Massari, 2023). Between 2015 and 2018, an estimated 1300 irregular travellers were charged as human smugglers in Italy, many of whom were simply forced to steer or navigate the boat they were travelling on (Patanè et al., 2020). The validity of Frontex’s claims that human smugglers are on board is hard to prove. The images make a simple determination of guilt based on the fact that a vessel with irregular travellers was on board. The images in Figure 5 reinforce assumptions of blanket suspicion and criminality for irregular travellers at sea.
Frontex blurs the line between victims and criminals. All boats and dinghies travelling irregularly are framed as threatening, always containing smugglers. This visual conflation is not to the benefit of irregular travellers, who may be prosecuted as human smugglers, but instead to Frontex’s benefit, as it exploits this visual ambiguity to create insecurity around all boats with mixed migration flows. In doing so, every detection and interception that Frontex makes of a vessel with mixed migration flows is a success, as it can claim to have arrested a smuggler on every boat. By presenting this ambiguity visually, Frontex uses X as a platform for visual governance, defining its own account ability and effectiveness. A montage of images included in Frontex’s posts on X (formerly Twitter) on detecting and capturing human smugglers (2015–2021). This montage was created using ImageJ correcting for varying image sizes.
Discussion
What do Frontex’s pictures of its policing at sea want from those who view them? While reflecting on this question, it is helpful to unpack an inherent defect underlying Frontex’s attempts to improve its legitimacy. Like other transnational policing agencies, Frontex is marked by a ‘deep asymmetry in the social distribution of its benefits and burdens’ (Loader, 2002: 298). What Loader is referring to here is how EU citizens may perceive certain benefits from Frontex’s policing (e.g., reduced drug smuggling or orderly migration), without experiencing any negative consequences from said policing (e.g., being pulled back to an unsafe country). The separation between who benefits and who is burdened by Frontex’s border policing contributes to the limited political will for holding transnational policing agencies like Frontex accountable (id.ibid). It may seem as though irregular travellers receive the benefits of search and rescue at sea, however, they also risk being labelled a human smuggler if it is found that they assisted with navigating or steering the boat. If they are rescued at sea by the Libyan coast guard they may also be detained and tortured.
The separation of the benefits and burdens of border policing means Frontex’s images at sea appeal to different audiences: irregular migrants and EU citizens. Frontex wants the pictures to deter irregular travellers from embarking on their journeys. The posts signal that the routes are dangerous and the vessels are unseaworthy. There is no guarantee that, once detected by Frontex, an irregular traveller will be disembarked to a place of safety or be safe from prosecution for human smuggling. The images and text portray effectiveness in the detection and interception of irregular travellers. The posts serve to trigger hesitancy and danger in the viewer who may consider travelling irregularly to the EU. At first glance, this may seem unreasonable, farfetched, and highly speculative that the images are attempting to deter irregular travel; however, Frontex is not the first European entity to use social media to try and prevent migration (Brekke and Thorbjørnsrud, 2018).
For EU audiences, Frontex’s images are meant to convey effectiveness in maintaining order at Europe’s sea borders. Frontex’s image work on X demonstrates account ability, which thereby promotes the legitimacy of the agency. Additionally, open waters are presented as vulnerable and a dangerous space where drug and human smuggling is rampant. Images portraying Frontex intercepting illicit goods and arresting smugglers help sustain the perception of the agency as essential to European security. On X, Frontex curates a perception of itself as a crime fighting agency that abides by its international human rights obligations. The photos of its officers racing to perform search and rescue are intended to generate empathy for the officers and the heroic and challenging work they do. The posts and images aim to evoke feelings of pride and effectiveness within the agency.
An unintended consequence of the routine display of irregular travel and crime is that it depicts Frontex as ineffective. Nevertheless, the repetition of these images also reifies the existence of the agency and its mandate. The endless stream of images depicting unseaworthy vessels and interception at sea, touch upon the maxim espoused by Sir Robert Mark, the former London Metropolitan Police Commissioner, who stated that ‘the real art of policing a free society or democracy is to win by appearing to lose; or, at least, to win by not appearing to win,’ (Mawby, 2013: 23). Mawby (2013: 24) explains this maxim as the delicate balance that the police must appear to perform, as they must simultaneously maintain order and sympathy from the public. Frontex’s image work on X leans into generating public sympathy for the agency. Frontex portrays itself as critical to maintaining the integrity of the EU’s external border, while simultaneously attempting to garner sympathy from the public for its heroism.
Finally, the analysis presented in this paper has several limitations that I will address here. Throughout the paper, I have assumed that Frontex is intentional about its social media use, deploying a long-term strategy to shift its portrayal from that of a search and rescue actor to a law enforcement actor. The posts on X may not be part of a long-term strategy, but rather a series of coincidental, haphazard decisions from Frontex’s press office. Nevertheless, if this is the case, it does not remove the fact that the accumulative effect—whether intentional or not—of posts that display heroism, overcrowded boats, and trophy shots can present narrow understandings of what Frontex ought to be accountable for, and omit some of its responsibilities to respect the human rights of irregular travellers.
Additionally, when creating the montages, I chose images that were indicative of the theme and sub-themes, as well as piqued my interest (i.e., punctum). The photos I selected were not necessarily wholly representative of how Frontex has portrayed itself visually, as I analysed one-fifth of the total images Frontex uploaded onto X between 2015 and 2021. Images are also inherently polysemic, and no one image can be interpreted in isolation (Barthes, 1977: 38–39). The way I saw and coded the posts may or may not be replicable. By coding and comparing images within a collection of similar photos, I aimed to be as transparent as possible about what I saw and did not see in Frontex’s images.
Conclusion
The agency’s image work on X means it is active in shaping its accountability, demonstrating where and how it is effective, and omitting details that might reflect how the agency neglects its human rights obligations. This paper reveals that Frontex’s visual accountability on digital platforms demonstrates how visual governance operates through images as both critical inputs and deliberate outputs (Massari, in this special issue). Images function as inputs by shaping public perception and discourse on migration. This was demonstrated through Frontex’s visuals of unseaworthy vessels that justify intervention. Concurrently, images serve as strategic outputs, deliberately deployed by actors like Frontex to communicate, legitimize, and control narratives. For instance, when Frontex shares crediting images that depict heroic police trophy shots of drug busts, which enhance its perceived effectiveness and garner public sympathy. Frontex shares images on X to define its own terms of accountability, often by omitting details like safe disembarkation or by visually conflating irregular travellers with criminals. This illustrates an iterative relationship where visuals simultaneously inform and are shaped by governance structures, actively influencing policymaking and the public understanding of border management.
Frontex uses visuals to define its responsibilities and establish criteria for evaluating its effectiveness and accountability. Future research should investigate how border policing agencies use video to construct and manage narratives of accountability—especially as Frontex increases its output on platforms such as YouTube and Instagram.
Footnotes
Acknowledgements
I would like to thank Alice Massari and Anna Triandafyllidou for their helpful feedback.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.