Cloud and Kurosawa's Quiet Fury at the Platform Economy

Kiyoshi Kurosawa’s Cloud opened in Japan in September 2024 and reached a broader international audience through a slow distribution pattern across late 2024 and the first quarter of 2025. It was Japan’s official submission for the international feature Oscar and did not make the shortlist. The film is not the one people who have followed Kurosawa since Cure and Pulse might have expected, and it is better, I think, than most of the reviews registered in the first pass. It is the film in which Kurosawa stops using ghosts and starts using logistics.

The premise is simple enough to summarise without collapsing it. Masaki Suda plays Yoshii, a Tokyo-based reseller who buys stock cheaply (counterfeit goods, deadstock, the grey edge of the consumer supply chain) and flips it through an online marketplace at significant markups. The film’s first act watches him accumulate enough capital to leave his factory job and relocate with his girlfriend to a rented cabin outside the city. The second and third acts watch his customers, collectively, find him.

The platform as ghost

Kurosawa has spent four decades making films about the specific ways dread enters ordinary rooms. His early horror cinema (Cure in 1997, Pulse in 2001, Retribution in 2006) used the ghost as a structural device for something in Japanese modernity he was trying to name. The ghost was the displaced worker, the disappearing social fabric, the specific isolation of the late-bubble salaryman. Cloud does something I did not expect from him. It removes the ghost. The dread is now made of actual people, and the people are customers.

Yoshii’s buyers are not, in the aggregate, monstrous. They are specifically aggrieved, in ways the film takes seriously. One has been sold a fake luxury watch. Another has lost significant money on counterfeit medical goods. A third, whose thread the film follows most carefully, has been sold a piece of resale equipment that failed catastrophically in a way that cost him his own small business. These are, individually, minor commercial grievances. Collectively, through the coordination layer the platform provides, they become something else. The platform is what organises their fury into a physical presence.

This is the argument the film is making that I have not seen him make before. The coordination is the horror. The individual complaints are ordinary. The aggregation of the complaints, through the affordance of the digital platform that made Yoshii’s business possible in the first place, is what turns a reseller’s bad month into a mountain siege.

Suda doing the quiet work

Masaki Suda is a specific casting choice. He has been one of the more interesting Japanese leading men of the last decade, working across genres with a deliberately non-heroic screen register. His Yoshii is not a sympathetic figure in the conventional sense, and the film refuses to soften him. What Suda gives the role is a specific kind of affective flatness. Yoshii is not a villain. He is a man running a small business within the rules of the platform that employs him. The flatness is the performance, and Suda sustains it across the runtime without letting it tip into caricature.

The scene I keep returning to is a negotiation early in the first act. Yoshii is buying a pallet of counterfeit merchandise from a supplier who is himself the middle link in a longer chain. The scene is shot in a warehouse in flat fluorescent light. The dialogue is about margins, turnover rates, the specific resale ceilings on particular categories. Suda plays Yoshii as a junior analyst at a company he is not consciously aware he works for. The company is the platform. The platform is everywhere and nowhere, and Suda’s Yoshii is a man who has stopped asking where his employer actually is.

The second act, and the shift

The film’s second act is the shift most reviewers flagged, and some did not forgive. Yoshii moves to the cabin, his customers begin to close in, and Kurosawa’s tonal register pivots from procedural to something closer to his older genre work. The shift is abrupt. I think it is deliberate.

The cabin sequences are shot in a specific cold palette (grey light, muted browns, an absence of the warm sodium tones of the earlier Tokyo scenes) and the camera is more static. Kurosawa is moving his visual grammar closer to the genre pieces he made in the 2000s, but the antagonists are now, crucially, not ghosts. They are men with backpacks, driving rental cars up a mountain road, coordinating over a messaging app.

The action choreography, when it arrives, is staged with a specific unheroic flatness. A shootout in the second half is not the bravura set-piece the premise suggests. It is a confused, embarrassing, mostly badly-aimed sequence in which amateurs try to do things they have seen in films and fail at them in specific ways. Kurosawa is not celebrating the violence. He is showing what happens when the platform teaches a dozen people to converge on a target and the target happens to own a firearm.

The girlfriend, and the film’s ethical centre

The character I want to mark is Akiko, Yoshii’s girlfriend, played by Kotone Furukawa. Furukawa’s performance is the film’s moral counterweight. Akiko is not a warning figure in the conventional sense; the film does not give her the speech that says stop doing what you are doing. What she does instead is watch Yoshii across the first act with a specific and increasing distance, and the distance is the film’s most sustained piece of emotional writing.

There is a scene, roughly forty minutes in, where Yoshii explains the margins on a transaction and Akiko asks a specific clarifying question about the product’s provenance. Yoshii answers. Furukawa lets a beat pass. The beat is the scene. The film, I think, is in that beat. Akiko has registered, before Yoshii has, that the job he is doing has a cost he is not calculating. The rest of the film is waiting for Yoshii to catch up.

The coordination, and what it names

What Kurosawa is doing at the level of the film’s larger argument is something I have not seen in Japanese cinema at this register before. He is taking the specific shape of the 2020s platform economy (the reseller, the aggrieved customer, the review system, the group chat, the logistics of converging on a physical location) and treating it as a horror object. The horror is not the individual platforms. It is the emergent behaviour the platforms make possible.

This is a specific kind of political filmmaking that refuses to name its politics. The film never says anything editorial about Amazon, Mercari, Yahoo Auctions, or any specific marketplace. It does not have to. The behaviour it depicts is the behaviour any viewer who has used any of these services will recognise at the structural level. The rage the customers direct at Yoshii is rage I have felt at a bad seller, expressed in a form I am grateful never to have felt.

What the film refuses

Cloud refuses several things conventional thrillers do not. It refuses to make Yoshii a hero in his own survival. It refuses to let the customers be a mob in the reductive sense. It refuses a tidy third-act revenge resolution. And it refuses, most importantly, the consolation of saying that the world depicted is a dystopian variant of the one we live in. The world is the one we live in. The horror is that the film did not have to invent anything structural to stage it.

The production and the register

The film is a Nikkatsu / Tokyo Theatres production with a reported budget in the low-single-digit-million-yen-billion range (numbers I would hedge on: Kurosawa has spoken in interviews about the production being efficient rather than lavish). The photography is by Yasuyuki Sasaki, whose work across Kurosawa’s recent films has been quietly excellent. The score, by Yusuke Hayashi, is restrained and mostly absent in the cabin sequences. The sound design, by Masayuki Iwakura, is the film’s most continuous craft element and deserves specific mention: the ambient hum of the Tokyo sequences is audibly different from the specific cold silence of the mountain, and the difference is doing real work.

What stays

Cloud is not Kurosawa’s most accomplished film. Cure remains that. What it is, and what I think it will be remembered as, is the film in which his specific cinema of dread adapted itself to the economy we are actually living inside, without compromising the register he has built across three decades. Watch it on streaming when it reaches the territory you live in. Watch it late at night. Watch it after you have packaged a parcel for dispatch. The film will sit with you differently depending on which side of the transaction you have most recently been on.