By Javier Surasky
Why does a site devoted to analyzing artificial intelligence, power, and global governance include a section on films? One possible answer is that cinema helps make complex issues more accessible, but the decision goes further than that.
Films are spaces where societies create imaginaries, build shared narratives, normalize relationships—including power relations—and project possible futures. That is why I decided to create a section on AI films at Global Radar Analytics, aside from the fact that I have a cinephile side I have no intention of hiding.
I do not present films to discover what will happen with artificial intelligence or to reconstruct its history. Although there are excellent documentaries on the subject, here I choose other genres, better suited to asking questions than to looking for answers.
Film as a Window onto AI
Paula Murphy argues in AI in the Movies (2024) that studying cinematic representations of AI allows us to observe how we think about this technology, how we react to it, what excites us, what we fear, and how our relationship with imagined artificial intelligences also forces us to reconsider how we define ourselves as human.
When we watch Blade Runner, The Matrix, Her, or Ex Machina, the least interesting question is whether they accurately portray a future technology or reflect the scientific knowledge of their time. It is more interesting to ask what they understand by intelligence, what they consider properly human, who owns the technology, who makes decisions, what relationships of dependence it creates, what forms of exploitation appear, and what institutions exist in response to the technological developments they depict.
There is also a historical issue: AI “in the movies” and AI as it actually exists have not developed at the same pace or in the same direction. Murphy (2024) shows that, since the 1950s, cinema has portrayed strong artificial intelligences endowed with consciousness, emotions, motivations, and capabilities that have still not been achieved.
Many of the images through which we think about intelligent machines were created before today’s generative systems entered our everyday lives. Cinema confronted us with conscious robots, rebellious machines, intelligences superior to humans, omnipresent surveillance systems, artificial companions, and blurred boundaries between organisms and machines. Bory, Natale, and Katzenbach (2025) distinguish between strong AI narratives, centered on future technologies capable of approaching or surpassing human intellectual capacities, and weak AI narratives, concerned with the functioning and consequences of the technologies we actually use.
While science fiction has favored the former, the latter are less spectacular but much closer to several of the problems we face today.
The distinction is useful because a story about a conscious machine threatening humanity can help us think about agency, autonomy, or control, but it can also draw our attention away from less cinematic yet more urgent issues: algorithmic systems that classify people, concentrate decision-making power, monitor workers, mediate social relations, or distribute benefits and costs unequally.
Films Build AI Imaginaries
Based on an analysis of 300 science-fiction films and a study of how different audiences perceive them, Elena Denia (2025) proposes thinking about AI narratives through three coordinates: apocalypse, assistance, and transcendence. Stories can present AI as a threat, as a tool for improving human capacities, or as part of a profound transformation of our own condition. This allows us to move beyond the opposition between utopia and dystopia and recognize multiple shades of gray and possible interpretations. After all, we have been debating the ending of 2001: A Space Odyssey for more than half a century, and not even Kubrick’s own explanation of its meaning was enough to settle the debate.
Denia also showed that audiences with less proximity to AI tend to lean more toward apocalyptic narratives, while those with greater knowledge of or interaction with the technology develop more nuanced positions. This finding shows that stories and social perceptions influence one another: the narratives surrounding AI affect the way a society receives a technology and can shape expectations about its development and regulation, offering us categories, characters, threats, and futures through which a difficult-to-understand technology becomes easier to grasp.
After analyzing 113 films, Peiro, Loup, and Aroles (2025) reach a compatible conclusion by identifying four major forms of AI representation: AI as individual support for human beings, as an individual antagonist, as a system created to help people collectively, or as an institutional system set against them. In every case, they find that these representations have a symbolic and political dimension that contributes to shaping social perceptions of AI.
This opens up a possibility that the Global Radar Analytics film guides incorporate: when a film gives AI a prominent role, it is worth asking whether it presents it as a tool, a subject, a companion, an authority, a commodity, or a threat; whether someone controls it or it exercises power over others; whether it is concentrated, distributed, or embodied; and who is exposed to its decisions. Just as important is asking which decisions and actors remain off camera.
Imagining Futures Means Imagining Ways of Governing Them
Macintosh (2025) analyzes how persistent representations of rogue artificial intelligences in film and television can contribute to reinforcing negative views of AI.
The problem is not that fiction departs from technical reality—after all, it is fiction—but this does force us to ask what happens when the images that circulate most widely are those of conscious machines attacking their creators, while the concrete impacts of actually existing systems are pushed aside. In other words, cinema helps us identify which problems become socially visible and which do not.
It would also be a mistake to speak of “AI cinema” as if it expressed a universal perspective. Anne Burkhardt (2026) shows how strongly the Global North has dominated both the literature and cinema of science fiction and artificial intelligence. In her research with Latin American filmmakers, other concerns emerged that receive less attention in AI cinema, such as labor exploitation within technological production chains or data colonialism.
Because imaginaries are produced from specific historical, social, and political contexts, thinking about the place from which a film is told—understanding “place” in a sense that goes well beyond territory—becomes central. The same technology can appear as a promise of efficiency from one position and as a tool of control from another.
Valdivia Alonso (2026) proposes viewing popular cinema as a symbolic infrastructure in which imaginaries are configured and sociotechnical futures are rehearsed. From this perspective, films confront us not only with questions about what technologies might exist, but also about who will have authority over them, under what rules they will operate, what forms of violence will be accepted, and what relationships between humans and intelligent systems will come to appear normal. A cinematic future can therefore be read as a hypothetical scenario of technological governance, because when a film presents an AI administering care or making security decisions, it is also distributing capacities and authority within the world it constructs.
Cinema helps make some futures plausible and others difficult to imagine, thereby shaping the horizons within which technologies are received and governed (Valdivia Alonso, 2026), and in doing so it influences the social acceptance of some technologies and the rejection of others.
Watching Is Not Enough
Showing a film does not by itself produce critical reflection. We can perfectly well watch a movie for entertainment or analyze it from a purely artistic perspective, but if we want to use it to think about AI, we need some kind of structure that helps us stop and consider aspects the narrative may move past quickly.
This is where the Global Radar Analytics movie guides come in. They do not seek to tell viewers what a work means or offer a “correct interpretation.” Instead, they provide an entry point for making connections between the cinematic experience and issues related to artificial intelligence.
Each guide identifies a general theme, proposes an analytical framework, selects relevant scenes, introduces questions for discussion, and connects the cinematic narrative with political, social, cultural, or ethical issues related to AI.
The aim is to move away from asking what happens in the film toward questions that require more active participation from the viewer: From what position is the story being told? What idea of intelligence does it propose? What power relations appear? Who makes the decisions, and who bears their consequences? What alternatives remain outside the story? What does the film not tell us?
This basic approach means there is no “correct” way to use the guides.
In a class or discussion session, a guide can be consulted before watching the film to identify issues worth paying attention to. For someone organizing or leading a discussion, it can help select relevant scenes, decisions, or dialogue. After the screening, it can serve as a starting point for collective conversation and, in a broader piece of work, as a bridge toward introducing academic literature.
There are also situations in which showing an entire film is neither possible nor desirable. In those cases, having a small selection of scenes that allows a specific issue to be explored can be useful.
Films Can Also Be Debated
For me, when it comes to film and AI, good questions are usually more useful than any closed answer.
Two people can interpret the same scene differently, and both readings can be valuable if they are able to make their assumptions explicit and argue for or against them. The point of the discussion is not to find the “true” meaning of a film, but to make visible the different readings it allows and the elements on which each one rests.
That is why the guides are designed as tools for thinking with the film, from it, and, when necessary, against it.
We should remember that artificial intelligence raises questions whose development remains open, and cinema has been rehearsing imaginary responses for a long time. Even before there was a discipline called artificial intelligence, we already encountered artificial machines associated with labor, power, rebellion, and social transformation in films such as Fritz Lang’s Metropolis, released in 1927, with its female robot used to lead workers toward disaster.
Almost a century later, we can look back and observe the imaginaries, fears, and desires those stories presented, which ones survived and which disappeared. That journey allows us to question the decisions we are making today about the development, use, and governance of artificial intelligence—and the extent to which some of those decisions still rest, at least in part, on futures we first saw on screen.
References
Bory, P., Natale, S., & Katzenbach, C. (2025). Strong and weak AI narratives: An analytical framework. AI & Society, 40, 2107–2117. https://doi.org/10.1007/s00146-024-02087-8
Brauner, P., Glawe, F., Liehner, G. L., Vervier, L., & Ziefle, M. (2025). Mapping public perception of artificial intelligence: Expectations, risk–benefit tradeoffs, and value as determinants for societal acceptance. Technological Forecasting and Social Change, 220, 124304. https://doi.org/10.1016/j.techfore.2025.124304
Burkhardt, A. (2026). Artificial intelligence in the work and view of Latin American filmmakers: Insights from a qualitative interview study. AI & Society, 41, 969–987. https://doi.org/10.1007/s00146-025-02577-3
Denia, E. (2025). AI narratives model: Social perception of artificial intelligence. Technovation, 146, 103266. https://doi.org/10.1016/j.technovation.2025.103266
Macintosh, K. L. (2025). Rogue artificial intelligence, science fiction, and the law. Santa Clara High Technology Law Journal, 41(1).
Murphy, P. (2024). AI in the movies. Edinburgh University Press.
Peiro, M., Loup, P., & Aroles, J. (2025). The fictional archetypes of AI: For a qualitative-quantitative analysis of representations in films. M@n@gement, 28(3), 1–20. https://doi.org/10.37725/mgmt.2024.9763
Valdivia Alonso, D. (2026). Imagining AI futures in mainstream cinema: Socio-technical narratives and social imaginaries. AI & Society, 41, 6125–6135. https://doi.org/10.1007/s00146-026-02880-7