Rethinking AI Production: Why ‘cheap and fast’ isn’t always cost-effective for agencies

The pitch was irresistible: generative AI would democratise production, slash budgets, and deliver Hollywood-grade visuals at a fraction of the traditional costs. Gen-AI tools like Runway, Pika, and OpenAI's Sora (with the latest addition of the uber powerful Sora 2) flooded the market with promises of instant cinematic magic. Yet, a year into widespread adoption, creative agencies are confronting an uncomfortable truth: high-quality AI production isn't cheap, and the economics are far more complex than anyone anticipated.

What began as a cost-cutting revolution is morphing into a new line item on production budgets. The infrastructure costs alone are substantial: getting GPU clusters, cloud rendering farms, or enterprise-grade AI subscriptions doesn’t come cheap. Runway's Standard plan costs approximately $35 per month with 625 credits, while the Pro plan at $95 monthly offers unlimited generations. For Sora, access comes bundled with ChatGPT Pro at $200 per month, and Midjourney's plans range from $10 to $120 monthly depending on usage needs.

Then there's the human capital: AI operators who understand prompt engineering, post-production artists who can blend synthetic and real footage, and creative directors who can art-direct algorithms. The promise of savings hasn't vanished, but it's been postponed, contingent on mastering a pipeline that's still being written.

Read On: AI in creative advertising: Game-changer or legal nightmare?

Neville Shah, Chief Creative Officer at FCB Kinnect, cuts through the hype with characteristic directness, saying, "AI is a tool, and I think everyone understands that. There is a flurry of 'I want AI content, I want AI content,' but as soon as you go to get AI content, their legal teams don't allow it. If you're a multinational, your legal team will not allow it, because the laws around it are so not done yet that everyone is a little wary."

The regulatory vacuum isn't just a theoretical concern. Brands are discovering that AI-generated content exists in a grey zone where copyright, liability, and authenticity intersect uncomfortably. For risk-averse legal departments, that's reason enough to pump the brakes.

But even when legal clearance comes through, the production reality is sobering. "A film that is being made in AI plus production is more expensive than a film that's getting made in AI alone or just in production. That blending only takes so much time, brainpower, and effort," Shah explains. Hybrid productions, where AI-generated elements are integrated with live-action footage, require meticulous attention to lighting consistency, motion tracking, and colour grading. The seams between real and synthetic must be invisible, and making them so is labour-intensive work that demands both technical expertise and creative judgment.

The Quality Gap Nobody Talks About

The democratisation narrative around AI video generation obscures a crucial distinction: the difference between ‘outputs’ and ‘deliverables’. Anyone can generate a passable AI video in minutes using consumer tools. But creating something that meets advertisement standards, that carries brand equity, that doesn't trigger the uncanny valley response in viewers.

"The other misnomer is that AI is cheap. Good AI is not cheap. And I don't just mean for the planet, it's also costly to do good AI work," Shah points out. "You can put up stuff that is one point-and-shoot, then you'll cut to it, and you'll have a consistency problem. Fixing all that costs money. It is possible, but it costs money." Challenges like character consistency across shots, maintaining spatial logic in generated environments, or even ensuring that brand assets appear correctly somehow multiply in AI production.

Read On: How ad agencies are customizing AI models

Traditional filmmaking does not pose these problems, as physical reality and continuity supervisors are obviously the strong points. AI production requires new methodologies, and agencies are building those workflows from scratch.

The learning curve is steep, but not a hopeless one. Shah notes a dramatic shift even within a single year: "Last year, we struggled for a one-minute film. This year, it's much easier. Still, the economics of it need constant conversation." The technology is evolving rapidly, but so is the understanding of how to deploy it effectively. Early adopters are discovering that AI production excellence requires the same kind of institutional knowledge that traditional production companies spent decades accumulating.

Siddhant Sethi, Senior Manager Founder's Office at White Rivers Media, frames the shift as a redefinition rather than a reduction of costs, saying, "AI video tools may promise affordability, but cinematic quality still comes at a creative cost. Generative AI has undeniably accelerated production workflows, yet achieving true broadcast or brand-ready quality demands more than algorithmic efficiency. The expense doesn't vanish, it just evolves. What once went into logistics and on-ground shoots now fuels GPU-heavy rendering, advanced post-production, and specialised AI operators who can fuse art with precision."

The Human Cost Equation

Beyond the technical and financial considerations lies a more uncomfortable question: the human cost of AI adoption. Shah doesn't shy away from it. "There's also a human cost. What about the person whose job you've taken?" The displacement anxiety in the production community is very much real. Roles of junior artists, background actors, and location scouts that once provided entry points into the industry are being automated. The long-term implications for talent pipelines and creative diversity remain largely uncertain.

Global award shows are simultaneously responding by drawing red lines. "Very famous award shows are now saying that if your film has X percent of AI, we will not allow it for a nomination. Why are they doing that? Because they're trying to protect people's jobs. You're generating actors through AI and doing all of that, so there are human and moral questions that need to be answered. The world hasn't answered them yet," Shah notes. These restrictions reflect broader industry tensions about authenticity, craft, and the value of human labour in creative work.

Read On: AI can cut costs but not replace emotions, asserts industry

For ad agencies navigating this landscape, the strategic question isn't whether to adopt AI, but how to deploy it thoughtfully. "For agencies, this shift means balancing speed with storytelling depth, ensuring that the output feels cinematic, not synthetic. In India, where content must marry scale with emotion, AI is less a shortcut and more a new craft discipline. True cost-effectiveness lies in mastering the technology, not merely adopting it," Sethi observes.

The most successful implementations are those that treat AI as one tool in a larger creative toolkit. Pitch decks, reference videos, or concept visualisations are areas where AI excels and delivers genuine efficiency gains. "You can make presentation boards, pitch decks, reference videos, or short 10–12 second content on AI. It's a great automation tool, and it will become part of the industry very soon," Shah says. But replacing entire production pipelines wholesale? That's where agencies are discovering the limits of the technology and the enduring value of human craft.

The Hybrid Creative Philosophy

Deepshikha Bhardwaj, National Lead of Media Strategy at Schbang, distills the current reality succinctly, saying, "AI has streamlined content workflows, but cinematic quality still requires significant human and technical investment. The costs haven't disappeared – they've shifted. Precision prompting, model training, post-production, and visual consistency now demand specialized expertise. In essence, AI has redefined the production stack, not reduced it."

This redefinition requires agencies to develop new competencies. The role of the ‘AI creative operator’ is emerging for someone who bridges technology and storytelling, or who can art-direct an algorithm as effectively as they can direct a cinematographer. These professionals command premium rates because their skill set is rare and valuable. Hiring them adds to overhead, even as it unlocks new creative possibilities.

Sethi emphasizes that the emotional resonance audiences expect from brand work still originates with human creators, saying, "AI can accelerate creation, but true storytelling still requires a human touch. Managing the balance between AI-driven efficiency and high-production storytelling now calls for a hybrid creative philosophy. Agencies are using AI for scale, automation, and rapid turnarounds. However, the emotional depth that audiences associate with strong brand narratives still comes from human craft." The process typically begins with AI-generated concepts that are later elevated through human intervention.

Shah is blunt about the creative limitations: "But this myth that it can become creative, if there is no empathy, how can it become creative?" Algorithms can pattern-match and recombine, but the spark of genuine creativity – the ability to understand human experience and translate it into compelling narrative – remains distinctly human. AI is powerful at execution, less so at ideation that matters emotionally.

Read On: Smarter, cheaper, faster: Has AI cut ad production costs by up to 60%?

The Way Forward

Looking forward, the optimism is cautious but real. Shah envisions a future where AI unlocks genuine production efficiencies: "In the future, I feel it will be more convenient because we'll have mastered certain techniques. For instance, I won't have to shoot in South Africa, as I could artificially generate South Africa. That's a possibility. Are we absolutely there yet? I don't think so." The destination is visible, but the journey is longer than the early hype suggested.

For now, agencies are in a transitional phase, investing heavily in both technology and talent to build AI-augmented production capabilities. The economics will likely improve as tools mature, as institutional knowledge accumulates, as workflows become standardized. But the notion that AI represents a simple cost substitution has been thoroughly debunked.

The real value proposition of AI in production isn't about replacing traditional methods wholly, it's about expanding what's creatively possible within existing budget constraints. A mid-sized agency can now attempt visual effects that once required major studio resources. A brand can test multiple creative directions quickly before committing to full production. These are genuine advantages, but they come with new costs and new complexities.

As Sethi frames it, "Ultimately, AI is a remarkable enabler, but the enduring magic of brand storytelling lies in human vision, not machine speed." The agencies winning in this new landscape aren't those chasing the lowest production costs, but those investing strategically in the hybrid future, where human creativity and artificial capability work in concert, each amplifying the other's strengths. That future is, without a doubt, expensive enough, but it might just be worth it.