Abstract Science fiction (SF) films reflect and shape sociotechnological realities as well as associated narratives and imaginaries. As such, they spark societal debates about the design of technological futures. Yet, the cinematic discourse on emerging technologies is heavily dominated by Western perspectives, while little is known about how these technologies are represented in SF films from the Souths. The present study addresses this gap by analysing a sample of ten Latin American (LatAm) films dealing with artificial intelligence (AI) - a key emerging technology of recent times - using qualitative film analysis methods and focusing on key representations and narrative motifs. These are interpreted in light of previous research on AI narratives and representations as well as region-specific contexts and debates. The analysis is further supported by findings from a previous study based on ten in-depth interviews with the directors of these films. The analysis reveals that AI in LatAm films is portrayed in markedly different ways from Western SF, with a focus on sociotechnological issues such as the unequal, colonial power structures underpinning AI, and a preference for depictions of ‘weak AI’ over ‘strong AI’. Consistent with this, AI is rarely portrayed as anthropomorphic, human-like, or agentic, instead serving as a tool of oppression wielded by authoritarian governments or tech companies. However, most of the films also present strategies of empowerment and resistance to these threats, linking to Latin America’s long tradition of creative and intellectual opposition to imported Western worldviews, as well as its engagement in debates on the decolonization of technology. The findings emphasise the films’ potential to stimulate decolonial practices of resistance - e.g., by subverting rather than reproducing Western SF narratives - and to foster meaningful debates on AI by centring human empowerment in the face of AI-related hazards instead of fuelling fears of AI extinguishing humanity.