Synthesia CEO Calls for Video Production Rethink as AI Avatars Scale Enterprise Training

AI video platform targets corporate learning budgets with synthetic presenters, raising questions about ROI versus traditional production

Victor Riparbelli, CEO of AI video generation platform Synthesia, is making an aggressive pitch to corporate finance teams: the traditional economics of video production no longer apply.

In a Financial Times interview published today, Riparbelli argued that finance leaders need to fundamentally reconsider how they budget for video content, particularly in training and internal communications. The London-based startup has built its business model around replacing human presenters with AI-generated avatars that can deliver scripted content in multiple languages without cameras, studios, or production crews.

The timing of Riparbelli's comments comes as CFOs face mounting pressure to demonstrate return on investment for AI spending. While generative AI tools have dominated technology budgets over the past 18 months, finance leaders are increasingly scrutinizing which applications deliver measurable productivity gains versus experimental deployments.

Synthesia's value proposition centers on eliminating the fixed costs of traditional video production—studio rentals, equipment, crew labor, and presenter fees—which can run into tens of thousands of dollars per finished minute for corporate content. The company's software allows users to generate videos by typing scripts, selecting AI avatars, and rendering the final product without filming.

For finance organizations, the use case extends beyond marketing into areas like compliance training, policy updates, and onboarding materials—content categories that require frequent updates and multilingual versions. Traditional video production creates a bottleneck: once filmed, updates require reshooting, while translation means either subtitles or expensive localized production.

The challenge for CFOs evaluating platforms like Synthesia lies in quantifying the trade-offs. AI-generated presenters eliminate production delays and localization costs, but they also eliminate the authenticity and emotional connection of human presenters—factors that impact training effectiveness and employee engagement, even if they're harder to measure on a spreadsheet.

Riparbelli's "forget everything you know" framing suggests the company is encountering resistance from finance teams applying traditional video production cost models to AI-generated alternatives. The pitch appears designed to reframe the conversation: rather than comparing per-video costs, Synthesia wants buyers to consider the total cost of maintaining a video content library over time.

This represents a familiar pattern in enterprise AI sales—vendors arguing that legacy mental models undervalue their technology's benefits. For finance leaders, the question becomes whether Synthesia's approach represents genuine cost transformation or simply shifts expenses from production to software licensing while introducing new risks around content quality and employee reception.

The broader implication for CFOs: as AI tools proliferate across corporate functions, finance teams will need frameworks for evaluating not just upfront costs but the second-order effects on productivity, quality, and organizational culture. A cheaper training video matters less if employees disengage from AI presenters or if content requires more frequent updates to compensate for reduced impact.

What remains unclear from Riparbelli's comments is whether Synthesia has quantified these trade-offs with customer data—the kind of ROI analysis that would help finance leaders move beyond "forget what you know" and toward "here's what we've proven."