In a guest column for TheWrap, Albert Lai, global strategic director for media and technology at Google Cloud, argues that the most consequential use of artificial intelligence in entertainment is unfolding in back-office operations rather than in on-screen content. He sets the case against a period of financial strain, writing that production costs are climbing while subscriber growth has stalled and traditional advertising revenue keeps sliding. In that environment, he contends, studios are looking to operations for savings that content alone cannot provide.
Lai draws a distinction between single-purpose tools and coordinated systems. Point solutions can deliver quick wins, he writes, but they leave the underlying workflows fragmented. The larger returns, in his argument, come from what he calls agentic AI, meaning systems that link tasks across departments and outside vendors so that work moves through the pipeline with less manual handoff. He presents this coordination, rather than any individual feature, as the change that matters for the bottom line.
He supports the claim with recent examples. According to Lai, Warner Bros. Discovery cut the time needed to caption unscripted streaming content by up to 80 percent and roughly halved the associated costs using generative tools. He also states that Netflix applied generative AI across about 300 titles in the first half of 2026, concentrated in post-production processes. He offers these figures as evidence that the technology is already embedded in routine studio work rather than confined to experiments.
The main obstacle, in Lai’s view, is structural. Decades of mergers, he writes, have left studios with fragmented systems and data scattered across separate departments and external partners, which makes coordinated automation hard to achieve. He points to the Emtek Group as an example of a company folding automated promo generation, live sports analysis, and localization into a single framework he describes as a studio of the future, an attempt to replace scattered tools with one connected operation.
Lai closes on the plumbing that would make such coordination possible. He cites emerging standards, including the Model Context Protocol and agent-to-agent protocols, as ways for systems from different companies to work together securely. His conclusion is that the visible uses of AI in film and television draw the most attention, while the durable advantage will belong to studios that rebuild their internal machinery. Readers may weigh that argument against his role at a cloud provider that stands to benefit from the shift he describes.
