Canal+ has signed a multi-year partnership with Google Cloud that will give the French broadcaster’s production partners access to Veo3, Google’s generative video model, for use in pre-visualisation and historical reconstruction. Deployment begins in June 2026 across European and African markets.
The deal makes Canal+ one of the first legacy broadcasters to embed a frontier generative video tool directly into its production relationships. This isn’t an internal experiment or an innovation lab announcement. It’s a major European media group telling the production companies it finances that AI-generated video is now available as a working tool – and that it expects them to use it.
The immediate applications are specific rather than sweeping. Production teams working on Canal+-backed projects will be able to use Veo3 to pre-visualise complex scenes before committing to a shoot, and to recreate historical settings or moments from archival photographs. Canal+ says early pilot projects have cut pre-production cycles by as much as 30 per cent. If that figure holds at scale, the implications for production budgets and development timelines are substantial – particularly for the kind of ambitious period drama and documentary content that Canal+ and its subsidiary StudioCanal have built their reputation on.
Canal+ has been careful with the framing. Stéphane Baumier, the group’s CTO, has positioned the tools as creative enablers rather than replacements, describing them as a way to produce scenes that would be impossible using traditional methods. Production partners retain full editorial and creative control. The use of Veo3 is optional — available to those who want it, not imposed on those who don’t. Rights and asset ownership remain protected within a secure technical environment provided by Google Cloud, with assets encrypted on the platform.
The careful language is telling. Canal+ operates in a market where production talent is unionised, where the French film industry carries significant cultural and political weight, and where the EU AI Act is imposing new requirements around watermarking synthetic media and data handling. Getting the positioning wrong on generative video wouldn’t just be a PR problem – it could jeopardise the production relationships that underpin the group’s entire content strategy.
The Google Cloud agreement extends well beyond production. Canal+ will use Google’s AI infrastructure to accelerate video indexing across its content library, building a multimodal database that classifies content by sound, video and text. That database feeds directly into the Canal+ App’s recommendation engine, which is also being overhauled through a separate partnership with OpenAI announced on the same day. Google handles the indexing and classification. OpenAI powers the subscriber-facing search and discovery layer. Together they represent a comprehensive technology bet by a broadcaster that, until its demerger from Vivendi in late 2024, was not widely regarded as a technology-first company.
The strategic context matters. Canal+ listed on the London Stock Exchange in December 2024 and now oversees more than 40 million subscribers across 70-plus countries, following its acquisition of South African pay-TV operator MultiChoice last September. CEO Maxime Saada has set a target of 100 million subscribers by 2030 – a number that requires both subscriber growth in Africa and a materially better product in Europe. The AI partnerships are a direct response to that challenge: better content discovery to retain subscribers, and more efficient production workflows to control costs while maintaining output quality.
For the broader television industry, the significance of this deal is less about Canal+ specifically and more about what it signals. Major broadcasters have spent years talking about AI in cautious, exploratory terms – pilot projects, ethical frameworks, internal working groups. Canal+ has skipped that phase entirely and gone straight to deploying a frontier generative video model through its production supply chain.
The question now is whether other broadcasters follow. The BBC, Bayerischer Rundfunk and Sveriges Radio have published ethical guidelines for working with general-purpose AI. Netflix and several European public service broadcasters use AI-driven analytics tools for content measurement. But none has yet announced anything comparable to giving production partners direct access to a generative video model for use in commissioned content.
McKinsey has estimated that AI could reduce overall production budgets by 5 to 10 per cent, with as much as 20 per cent of original content spending shifting toward AI tools within five years. Canal+ appears to be betting that those numbers are real, and that the competitive advantage goes to whoever moves first.
Whether the rest of the industry agrees – or whether it watches from a cautious distance while a French broadcaster runs the experiment for them – will be one of the more revealing dynamics of the next 12 months.
