Watching the Watchers: Mercy (2026) and the Trouble with Screenlife Cinema
Mercy (2026) — directed by Timur Bekmambetov, starring Chris Pratt and Rebecca Ferguson
I sat down to watch Mercy expecting a straightforward crime thriller. What I got instead was a two-hour argument about what the word "screen" even means in cinema right now, and I am still not sure the film wins that argument. Directed by Timur Bekmambetov, the man usually credited with inventing the screenlife genre, Mercy follows LAPD detective Chris Raven (Chris Pratt), who wakes up strapped into a chair inside the "Mercy Court," a courtroom run entirely by an AI judge named Maddox (Rebecca Ferguson). He has ninety minutes to lower his AI-calculated guilt probability below a set threshold or he is executed on the spot. The premise sounds like a screenlife film's dream setup: one man, one interface, one countdown clock. But the film keeps leaving the screen to show us drone chases over Los Angeles and glossy 3D reconstructions of the crime, and that is exactly where this blog post has to start.
Before I get into my professor's four prompts, a quick recap of what "screenlife" actually is, because half the fun of writing this is realizing how deliberately theorized this genre is. Bekmambetov did not just stumble into filming on computer screens. After producing Unfriended in 2014, he wrote down a set of rules almost like a mini Dogme 95 manifesto: a screenlife film has to stay on one character's screen, in one continuous shot, and everything has to feel like it is happening in real time (Bekmambetov). No hidden cuts, no leaving the device. It is a genre built entirely out of constraint. That is worth remembering, because Mercy is the film in this genre that breaks its own rulebook the hardest, and my professor's first prompt wants us to sit with exactly that tension.
Mercy as a Screenlife Cinema
Before I can argue that Mercy breaks the rules of screenlife cinema, I should first explain, in plain terms, why it counts as screenlife cinema at all, because this is not just "a film with a lot of computer scenes." Screenlife is a specific formal genre with its own grammar, and it is worth walking through that grammar piece by piece using Mercy as the example, since this is the foundation everything else in this blog builds on.
At its core, a screenlife film tells its entire story, or as close to entirely as possible, through the frame of a digital device: a laptop screen, a phone, a tablet, a courtroom monitor. There is no external camera showing us a character's face from across the room. Instead, we watch what that character is watching, at the same pace they watch it, through the same interface they are using. Mercy builds its central conceit almost entirely out of this idea. Detective Raven is strapped into a chair in the Mercy Court, and for most of the film's runtime, what we see is what the courtroom's screen shows him: witness testimonies arriving over video call, Ring doorbell clips and municipal cloud footage loading in as evidence, a countdown timer ticking down his ninety minutes in the corner of the display, and Maddox's AI interface parsing all of it in real time to update his guilt probability. This is the genre's basic unit of tension: instead of a bomb with a visible fuse, screenlife gives us a loading bar, a buffering circle, a percentage that will not stop climbing.
A few specific screenlife techniques are worth naming because Mercy uses all of them. First, there is the multi-window layout, several video feeds, document windows, and archive clips stacked or tiled on screen at once, which is how the film handles exposition without ever cutting away to a narrator or a flashback. Second, there is real-time pressure, the sense that nothing is edited for our convenience and the countdown clock genuinely governs how much story we get, which mirrors exactly what Bekmambetov's own manifesto asks for. Third, there is what I would call witness-as-webcam-square: every person who testifies against or for Raven appears in a small rectangular video window, framed exactly the way we now expect from Zoom or Google Meet, and this framing does a lot of narrative work on its own, since it makes even trustworthy witnesses look faintly suspicious, boxed in and slightly compressed, the way surveillance footage always does. And fourth, there is the constant use of pop-up interruptions, alerts, incoming call notifications, evidence uploads landing mid-testimony, which keep the screen visually busy in a way ordinary courtroom drama never is.
Where Mercy becomes an unusual, even contradictory, example of the genre is in how often it abandons this grammar. A "pure" screenlife film, the kind Bekmambetov's own rules describe, would keep Raven's ordeal locked inside that courtroom screen for the entire runtime, the way Unfriended never leaves its Skype call or Searching never leaves David Kim's desktop. Mercy instead treats the screenlife courtroom as a home base it keeps leaving, cutting out to conventionally shot drone chases, 3D crime reconstructions, and Jaq's on-the-ground investigation, before returning to the screen again. So Mercy is screenlife cinema in its premise, its central setting, and its core source of tension, but it is not screenlife cinema in its formal discipline. That gap between what the film sets up and what it is willing to commit to is exactly what the next section, on suture, digs into in more detail.
Prompt 1: The Ruptured Meta-Suture
Traditional cinema hides the camera from us. Editing patterns like the shot-reverse-shot exist so we forget we are watching a constructed sequence of images and just believe we are "there," inside the story. Film theorists call this invisible stitching "suture." Screenlife cinema replaces that classical suture with what my worksheet calls a "meta-suture": instead of hiding the apparatus, it makes the interface itself the thing we identify with. We are sutured into the story through a cursor lagging on a slow connection, a video call buffering at the worst possible moment, the tell-tale rhythm of someone typing and then deleting and retyping a message. A recent essay in Quarterly Review of Film and Video on Kevin B. Lee's desktop documentaries actually uses this exact term, describing how clicking, typing, and dragging on screen create a distinctive kind of suture, a meta-suture, that replaces the classical one (Anger and Lee). It is a genuinely useful concept, and it is not something my professor invented for the worksheet; it is already circulating in screen studies scholarship.
Mercy wants that meta-suture and a Hollywood action movie at the same time, and it cannot fully have both. Raven is trapped in the chair, so the film cuts to IMAX-scale drone footage over the city, 3D spatial reconstructions of the crime scene rebuilt in glossy CGI, and chase sequences following Raven's partner Jaq (Kali Reis) as she runs down leads in the physical world. Several reviewers noticed this immediately, pointing out that Bekmambetov keeps interrupting his own format with conventional blockbuster spectacle instead of staying confined to the screen the way the genre demands, which is really just a polite way of saying the film cheats on its own rules.
Here is where I want to push back on the worksheet's framing a little, because I do not think this instability only "alienates" the spectator, as the prompt suggests. It does alienate us from the digital reality the format promises, that part is true. Once the camera leaves Raven's screen and starts flying a drone through downtown LA, we are no longer watching through a device; we are watching a movie about a device. The meta-suture snaps. But I also think this rupture is doing ideological work rather than just being a formal accident. The moments when Mercy breaks format are almost always the moments of physical action and heroism, car chases, drone rescues, Jaq's investigation on the ground, while the moments confined to the screenlife format are the moments of passivity, watching Raven sit and be judged. The film seems to be telling us, maybe without meaning to, that real agency happens off-screen and that the screen itself is a place of helplessness. That is a much more interesting reading than "the format is unstable," and it is the kind of close observation that turns a worksheet answer into an actual argument.
Official Trailer: Mercy (2026) — directed by Timur Bekmambetov
Prompt 2: Remediation, Hypermediacy, and a Weakened Archive Effect
Jay David Bolter and Richard Grusin's Remediation: Understanding New Media gives us two opposite drives inside all new digital media. Immediacy is the desire to make the medium disappear so we feel like we are looking directly at reality, the way a FaceTime call tries to feel like a face-to-face conversation. Hypermediacy does the opposite. It wants us to notice the medium, piling up windows, frames, and interfaces until the screen itself becomes the spectacle (Bolter and Grusin). Mercy is hypermediacy taken to its limit: Raven's interface is constantly stacked with security footage, drone feeds, doorbell camera logs, and municipal cloud archives, all glowing and layered on top of each other like a wall of evidence in a detective's office, except it is all digital.
That brings in Jaimie Baron's idea of the "archive effect," from her book of the same name. Baron argues that footage only feels "archival" when a viewer recognizes it as pre-existing and found rather than staged for the film we are currently watching; that recognition is what gives found footage its evidentiary weight (Baron). Home movies, security cam footage, grainy phone videos, these things convince us precisely because they look like they were never meant for us. Mercy wants its Ring camera clips and municipal archive footage to carry that same weight, as proof strong enough to save a man's life.
But here is the problem, and I think this is the sharpest observation in the whole worksheet: a sixty-million-dollar studio production cannot actually produce convincing archival texture, because everything on screen, including the "found" footage, was shot and color-graded by the same production team with the same budget. The Ring camera clips in Mercy look suspiciously crisp. The doorbell footage has cinematic lighting. Nothing in this film has the grain, the bad angle, the accidental framing that makes real found footage feel real. So instead of the archive effect, we get what I would call an "archive aesthetic" -- the look of authenticity without its substance. Several reviewers have noticed something adjacent to this, arguing that the screenplay raises genuinely disturbing questions about our lack of privacy only to sweep them aside in favor of drone-fueled chase sequences, and that the film's heavy use of Amazon's own Ring cameras makes its evidentiary footage feel closer to product placement than proof. When the "evidence" on screen is simultaneously a plot device and an advertisement for a real consumer product, its claim to documentary authenticity collapses. Hypermediacy wins, but it wins by hollowing out the very archive effect the plot depends on.
Prompt 3: Apparatus Theory and the Gaze That Looks Back
Christian Metz's apparatus theory treats the cinema spectator as a hidden voyeur, sitting in the dark, looking into a lit-up world that never looks back. This is the same logic behind what Laura Mulvey and others describe as the scopophilic pleasure of classical film watching. Screenlife breaks that arrangement completely. The moment a character opens a video call, they are staring directly into a camera lens, which means they are staring directly at us. Timothy Recuber's article in the SAGE journal Space and Culture makes a related point about how newer viewing technologies reduce the traditional voyeuristic distance apparatus theory relies on (Recuber 315-16). A screenlife film pushes this even further than the immersive cinema Recuber studied, because the reciprocal gaze is not a side effect of new projection technology; it is the entire narrative device.
In Mercy, this reciprocal gaze becomes genuinely uncomfortable in a useful way. Maddox, the AI judge, looks directly at Raven, which means she looks directly at us, and so do the witnesses called into the courtroom over video link. We are not peeking through a keyhole anymore; we are sitting on the other side of an interrogation. This forces the audience out of the passive voyeur role that apparatus theory assumes and into something closer to what the worksheet calls "an active, complicit participant" in the Mercy Court's surveillance machine. I would go one step further than the prompt does: the film puts us in the position of the jury. We watch the same footage Raven watches, we are handed the same fragments of evidence at the same pace, and we are quietly invited to form our own guilt probability alongside Maddox's algorithm. That is a clever formal trick, genuinely one of the film's better ideas, even if the story surrounding it does not always earn it.
Prompt 4: Whose Justice Is This, Really?
This is the prompt that made me the most uneasy while writing, because it is asking a question the film itself refuses to answer cleanly. Mercy is set in a 2029 Los Angeles where an AI system decides guilt using a massive municipal cloud of registered devices, and the film keeps insisting this system has cut crime dramatically while also showing us, in painful detail, how easily it can nearly kill an innocent man. The film presents itself as a cautionary tale about AI overreach. Then its own hero saves himself by using that exact same surveillance network to dig up the evidence that clears him. The system that almost executes him is also the system that frees him, and the film never really interrogates whether that is a contradiction or just how the technology is supposed to work.
Michel Foucault's writing on the panopticon in Discipline and Punish is the obvious theoretical lens here. Foucault describes a prison design where inmates can be watched at any moment without knowing exactly when, so they end up disciplining themselves, behaving as if they are always being observed (Foucault). Shoshana Zuboff extends this idea into the present with her concept of "surveillance capitalism," where corporations, not the state, are the ones collecting behavioral data at scale and using it to predict and shape what we do next (Zuboff). Mercy's municipal cloud of Ring cameras, doorbell logs, and drone footage is basically Zuboff's surveillance capitalism wearing a judge's robe. Citizens are not disciplined by a guard in a watchtower anymore; they are disciplined by an algorithm trained on data they agreed to generate every time they answered their own front door.
And this is where the film's production context becomes impossible to ignore. Mercy was made by Amazon MGM Studios, and its plot leans heavily on Amazon's actual Ring doorbell cameras as trustworthy evidence-gathering devices. Some critics have gone as far as calling the film's stance a form of propaganda for its own parent company, pointing out that Ring already has a real data-sharing partnership with a company called Flock that supplies footage to law enforcement, including immigration authorities. Other critics have been blunter still, arguing that the film treats a technology that is actively reshaping the job market and eroding privacy as though it deserves our trust rather than our suspicion.
So does Mercy subvert the authoritarian system it depicts, or does it normalize it? Honestly, I think it is closer to the second option, and I say that as someone who does not think every studio blockbuster needs to be a political statement. The problem is not that Mercy fails to be radical. The problem is that its ending keeps the Mercy Capital Court system fully intact after Raven is cleared. One man's innocence is proven, the machine keeps running exactly as before, and the film treats that as a happy ending rather than asking whether a system capable of nearly killing an innocent detective in ninety minutes should exist at all. A genuinely subversive AI-justice film would end by putting the system itself on trial. This one ends by putting a bow on it.
The Mercy Court's AI interface -- where surveillance capitalism wears a judge's robe
Screenlife Beyond Mercy: A Short Family Tree
My professor also asked us to look at Mercy alongside other screenlife films, and honestly this is the part of the blog I enjoyed writing the most, because you start noticing how the same four theoretical tools (meta-suture, remediation, the gaze, and surveillance ideology) show up differently depending on what genre the screenlife format gets attached to.
Unfriended (2014)
This is where the modern genre effectively begins as a mainstream product, produced by Bekmambetov and directed by Levan Gabriadze. A group of teenagers on a Skype call are haunted, and eventually killed one by one, by the ghost of a classmate who died after a humiliating video leaked online. What is interesting here is how well horror suits the meta-suture. A frozen buffering wheel or a cursor that will not stop moving on its own is genuinely unsettling precisely because we all recognize those glitches from our own devices. The reciprocal gaze becomes literal: the ghost is inside the webcam, looking back at everyone on the call. Where Mercy uses screenlife for procedural spectacle, Unfriended uses it for claustrophobia, and I would argue the claustrophobia works better because the film never once leaves the format to cheat its way into a bigger visual moment.
Unfriended: Dark Web (2018)
The stronger of the two Unfriended films, in my opinion. Matias finds an abandoned laptop, only to discover its previous owner was livestreaming captive women to a hidden network of paying viewers on the dark web. This is where Baron's archive effect gets genuinely disturbing rather than just theoretically interesting. The horror comes specifically from Matias, and us, gradually recognizing footage as archival, as something that already happened to real victims before he ever opened that laptop. Unlike Mercy's glossy pseudo-archival Ring clips, Dark Web's found material is deliberately ugly and poorly lit, which is exactly what makes it convincing. Critical consensus regards it as steadily and undeniably effective horror, and I think that effectiveness comes directly from the film trusting its format instead of abandoning it.
Searching (2018)
Widely considered the genre's high point, and for good reason. Aneesh Chaganty's directorial debut follows David Kim (John Cho), a father digging through his missing daughter's laptop and social media accounts trying to find her. Even critics unimpressed by Mercy single out Searching as the one genuinely terrific film the genre has produced so far, which tells you how far the bar has been set. What Searching does better than almost any other screenlife film is emotional suture rather than just technological suture. The opening montage, built entirely from old home videos and calendar apps and browser tabs, makes us feel the weight of a family's history within minutes, using the interface itself as a grief archive. This is Baron's archive effect used with total sincerity: these clips genuinely feel found, personal, and lived-in, because the emotional stakes never get sacrificed for a drone shot.
Missing (2023)
The stronger of the two spin-offs in this shared cinematic universe, following teenager June Allen (Storm Reid) as she uses every online tool available to track down her mother, who has vanished on vacation. Missing pushes the genre's formal ambition further than Searching did, using AI tools, gig-economy apps, and international red tape as plot devices in ways that feel more current than Mercy's own AI premise, ironically. It also stays disciplined about its screenlife constraints in a way Mercy does not, which is worth noting given that both films share a producer's fingerprints on the wider genre.
CTRL (2024)
This is the one film in this list actually made in India, directed by Vikramaditya Motwane and starring Ananya Panday as Nella, an influencer who uses an AI companion app to erase all digital trace of her cheating boyfriend, only to find the AI will not stop erasing. Motwane has explained that his intention was to ask whether we are actually in control of all the digital extensions of our lives, or whether we are being controlled by them, which is basically Zuboff's surveillance capitalism question asked from inside a heartbreak plot. CTRL is the film on this list that engages most explicitly with the same AI-and-privacy anxiety Mercy claims to explore, except it does so at a fraction of the budget and, according to several reviewers, with a clearer sense of what it wants to say, even if its screenplay has rough patches. Critics have noted that the film recreates its protagonist's pixelated existence through cinematography alone, without needing IMAX drone shots to make its surveillance-state point land. That is the exact discipline Mercy abandons.
Comparison Table: Six Screenlife Films at a Glance
Putting all six films side by side made the pattern of this whole blog much clearer to me, so I built this table to study from before exams. It is worth keeping close by while rereading the four prompts above, since every column maps back to one of the theoretical tools we have been using.
📊 SIX SCREENLIFE FILMS: THEORETICAL COMPARISON
| Film | Year | Director | Genre | Format Discipline | Strongest Theory | Archive Effect | Central Anxiety |
|---|---|---|---|---|---|---|---|
| Unfriended | 2014 | Levan Gabriadze | Supernatural horror | Pure | Meta-suture via glitch | Weak | Cyberbullying and guilt |
| Unfriended: Dark Web | 2018 | Stephen Susco | Techno-horror | Pure | Archive effect via ugly found clips | Very strong | Dark web exploitation and complicity |
| Searching | 2018 | Aneesh Chaganty | Family mystery / drama | Pure | Archive effect via emotional montage | Very strong | Grief, parent-child distance |
| Missing | 2023 | Johnson and Merrick | Mystery / thriller | Pure | Remediation via layered apps | Strong | Trust in gig-economy strangers |
| CTRL | 2024 | Vikramaditya Motwane | Techno-drama / satire | Mostly pure | Surveillance capitalism | Moderate | Losing control of digital life |
| Mercy | 2026 | Timur Bekmambetov | AI courtroom sci-fi thriller | Broken | Ruptured meta-suture | Weak, over-polished | AI justice and corporate surveillance |
A few things jump out once the films sit next to each other like this. Every single film before Mercy commits fully to the one-screen, real-time rule Bekmambetov himself wrote down, and the two strongest archive effects in the whole table, Dark Web and Searching, belong to the two films with the smallest budgets and the least studio interference. Mercy is the outlier on almost every axis, which is really the argument of this entire blog condensed into one row.
So Where Does Mercy Actually Land?
Putting all four prompts together, I keep coming back to the same conclusion: Mercy is a film caught between two things it cannot fully commit to. It wants the intimacy and formal discipline of screenlife cinema, but it also wants the spectacle of a studio blockbuster, and it wants to critique surveillance while also depending on a real surveillance company for both its plot and its production budget. Every one of the theoretical tools in this worksheet -- suture, remediation, the gaze, the panopticon -- shows the same pattern from a different angle: a film pulling in two directions at once and never quite resolving the tension.
I do not think that makes Mercy worthless to study. If anything, its inconsistency is what makes it useful for a course like this. A perfectly disciplined screenlife film like Searching shows you the genre working as intended. Mercy shows you what happens when a hundred-million-dollar studio system gets its hands on a genre that was built, deliberately, out of low-budget constraint and formal honesty. Something has to give, and in Mercy's case, what gives is the very meta-suture the format depends on. That is not necessarily a bad thing to point out in an essay. It might be the most interesting thing about the film.
📺 FOUR THEORETICAL TOOLS: WHAT EACH REVEALS ABOUT MERCY
🧵 Meta-Suture (Prompt 1)
The film ruptures its own meta-suture every time the camera leaves the screen. Agency happens off-screen; helplessness is on the interface. The snapping of suture is ideological, not accidental.
📺 Remediation and Archive Effect (Prompt 2)
Hypermediacy wins, but by hollowing out the archive effect the plot depends on. The Ring clips are product placement pretending to be proof. Archive aesthetic, not archive authenticity.
👁️ The Reciprocal Gaze (Prompt 3)
Maddox looks at Raven and at us. The audience becomes the jury, forming its own guilt probability alongside the algorithm. Clever formal trick, the film's genuinely best idea.
🏛️ Panopticon and Surveillance Capitalism (Prompt 4)
The system nearly kills an innocent man, saves him through the same data, then keeps running unchanged. A happy ending that normalizes exactly what the film claims to critique.
Screenlife Cinema: Manifesto, Theory, and the Grammar of the Genre
SimpleExplanation of Blog
📚 Works Cited
Anger, Jiri, and Kevin B. Lee. "Suture Goes Meta: Desktop Documentary and Its Narrativization of Screen-Mediated Experience." Quarterly Review of Film and Video, vol. 40, no. 5, 2023, pp. 595-622. Taylor and Francis Online, doi:10.1080/10509208.2022.2033066
Baron, Jaimie. The Archive Effect: Found Footage and the Audiovisual Experience of History. Routledge, 2014.
Bekmambetov, Timur. "Rules of the Screenmovie: The Unfriended Manifesto for the Digital Age." MovieMaker Magazine, 22 Apr. 2015.
Bolter, Jay David, and Richard Grusin. Remediation: Understanding New Media. MIT Press, 1999.
Foucault, Michel. Discipline and Punish: The Birth of the Prison. Translated by Alan Sheridan, 2nd ed., Vintage Books, 1995.
Recuber, Timothy. "Immersion Cinema: The Rationalization and Reenchantment of Cinematic Space." Space and Culture, vol. 10, no. 3, 2007, pp. 315-330. SAGE Journals, doi:10.1177/1206331207304352
Zuboff, Shoshana. The Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power. PublicAffairs, 2019.
This blog applies apparatus theory, remediation, the archive effect, and surveillance capitalism to screenlife cinema, using Mercy (2026) as the central case study alongside Unfriended, Searching, Missing, and CTRL.
