The Verge is reporting on the emergence of AI-generated films positioned to capitalize on the cultural moment around major theatrical releases, drawing a direct comparison to the direct-to-video knockoff industry that flourished in the era of physical media. The phenomenon represents a new and largely unregulated frontier in content production, where generative AI tools allow small operations to produce feature-length video at a fraction of the cost and time it would take a conventional film crew.
To understand why this matters, it helps to remember where the film industry has been before. The direct-to-video era of the 1980s and 1990s gave rise to studios whose entire business model depended on mimicking the titles and imagery of big theatrical releases close enough to confuse a distracted shopper. Companies like The Asylum became notorious for this practice, releasing what the industry calls mockbusters — films with names and cover art engineered to ride the marketing coattails of blockbusters without bearing any of the production costs. Sharknado aside, most of these films were cynical commercial products, and the audience for them understood exactly what they were getting. The model worked because distribution was cheap, confusion was easy, and there was always a floor of demand for something loosely resembling whatever was playing at the multiplex.
What generative AI has done is essentially remove the remaining friction from that process. Where a mockbuster still required camera operators, actors, locations, and editing suites, an AI-generated film can theoretically be assembled by a very small team using text-to-video tools, voice synthesis, and automated editing pipelines. The marginal cost of producing something that looks superficially like a movie has collapsed. And crucially, so has the time between a theatrical announcement and the ability to produce a response product. A film can now be generated fast enough to be ready when cultural interest in a given title peaks, rather than arriving months later when the moment has passed.
The timing around a release of the scale The Verge describes is significant. A Christopher Nolan film generates enormous pre-release conversation, the kind that floods social platforms with searches, trailer views, and recommendation queries. That search traffic is valuable real estate, and the likely reading is that AI slop films are being engineered not necessarily to be watched in any meaningful sense, but to capture placement in recommendation algorithms, streaming platform libraries, and casual browsing sessions where a thumbnail and a title are enough to trigger a confused click or an impulse rental. The revenue model is not theatrical — it is the monetization of ambient attention.
This creates a problem that is distinct from the one posed by traditional mockbusters, because the scale is different and the barrier to entry is lower than it has ever been. The Asylum needed staff, relationships with distributors, and some baseline level of production infrastructure. The new generation of AI content producers may need almost none of those things. That asymmetry is deeply uncomfortable for an industry already grappling with how to think about AI's role in screenwriting, visual effects, and voice acting. Here the technology is not augmenting human creative work — it is substituting for it wholesale, and doing so in a way designed to parasitize the marketing expenditure of studios that have spent tens or hundreds of millions making something genuine.
The consequences fall on several groups at once. For working filmmakers, editors, composers, and actors at the lower end of the production market — people who built careers doing exactly the kind of work that fills the middle and lower tiers of streaming libraries — this represents a direct and immediate competitive threat. For streaming platforms, the proliferation of AI-generated content creates a content moderation and curation problem they are not obviously equipped to handle. A library full of AI slop erodes the signal value of any given title and trains audiences to distrust recommendations. For audiences, the harm is subtler but real: time and sometimes money spent on content that was never intended to deliver anything beyond a click.
There is also a longer-term reputational question for the AI tools companies themselves. If the most visible consumer-facing use case for text-to-video generation turns out to be the industrial production of deceptive knockoff content, the political and regulatory pressure on those companies will intensify considerably.
What to watch for next is whether the major streaming platforms move to implement any form of AI-disclosure requirement or detection mechanism at the point of ingestion, before a title reaches a library. Some have begun experimenting with content provenance tools, but enforcement remains inconsistent. Also worth watching is whether any jurisdiction attempts to apply existing consumer protection or unfair competition law to AI-generated titles that are clearly designed to mislead. The technology has outpaced the rules before. It appears to be doing so again.
