Everything But the Cut

From the Moviola to Media Composer, from on-prem shared storage to remote post in the cloud, the way stories are cut has been reinvented time and again. But now, as AI and AI agents make their way into the post production world, some would have you believe professional editors are digging in their heels, reluctant to incorporate this new technology. At Avid, we’ve been part of numerous industry changes, so we set out to understand how true this was: were editors rebelling against the use of AI?

In partnership with ACE, we conducted research on the existing practices of using AI in post production, as well as the future of AI in media creation. 120 editors across the fields of film and TV responded to 17 questions about their feelings toward AI and where it belongs in post. TL;DR: 93 of the 120 — 77.5% — are already using AI in their work.

Without context, it might be easy to assume creators of any kind would resist AI, citing its limits on creativity, or questions around content ownership. However, it’s not just how many professional editors are using AI in their work, it’s how they’re using it that is the most intriguing.

Asked what they currently use AI for (editors could select more than one answer), 60.0% said transcription, 55.0% audio cleanup and 39.2% search.

When asked to name the single capability that would make the biggest difference to their work, 81.7% chose one of five that handle the work around the edit: audio cleanup, transcription and dialogue search, finding specific shots, metadata tagging or project organization. In our research, these same editors also clearly defined the one thing they're unwilling to hand over — authorship.

They're equally clear about how they want these AI tools delivered. 67.5% say it matters that AI is built into the tools they already use rather than accessed through a separate application, and 63.3% agree AI tools need tighter integration into existing workflows to be useful. Among daily and weekly users, both figures rise above 80%. They don't want another window to open; they want the environment they already work in to get materially smarter, allowing them more time to focus on what they do best: storytelling.

3 Big Ideas

Editors shared several ideas in our recent survey, so we’ve combed through the results for the 3 biggest:

• AI belongs in the dress rehearsal, never the final show: AI can be used to generate ideas, but it should never make the final cut.
• The need for speed: Editors who use AI regularly are likely saving time and investing it back into the work.
• More, or better: More time in the work means better results for what matters, the creative process.

The Dress Rehearsal

One editor mentioned using ElevenLabs for temp ADR, and Midjourney to illustrate potential pickup shots, then plainly stated: “I cannot see a way it would be useful for final outputs.” Another has written himself a personal rule — AI-generated material “will only be used for temp purposes, never in a final.” A third noted: “I don't want to generate an image. I want the AI to transcribe footage or find lines I'm looking for.”

AI is welcome in the dress rehearsal. It doesn't belong in the show.

That position is why the current market is a strange thing to watch. Generative images and video are used by 27.5% of these editors, well behind transcription, audio cleanup and search. Yet generation is precisely what most of this technology is marketed on. Meanwhile 86.7% agree AI is most valuable when it enhances rather than replaces creative intent — and the line is drawn hardest by the people who use AI most: 98% of daily and weekly users agree, compared with 78% of those who rarely or never use it. They want AI to buy back time for storytelling, not to do the storytelling.

The Need for Speed

Ask anyone in a leadership role at a studio and they’re likely thinking about AI in terms of efficiency and content optimization. So, we asked the editors: has AI helped you or your team work faster?

At first glance, the results were not conclusive, but if you break them apart by how often editors use AI, the data tells a more intriguing story. 76.2% of daily and weekly users say AI has made them faster, as do 60.9% of monthly users. Among those who use it rarely or never, that falls to 18.2%. Taken together, that's 70.8% of the 65 editors who use AI at least monthly. Nearly three-quarters of the "no noticeable difference" answers come from editors who use AI rarely or never, which is less a verdict on the technology than a description of what it means not to use something.

For those who rely on AI regularly, the time savings appear to be real. But with time saved, the next question is where that time goes.

Anyone who lived through the arrival of digital nonlinear editing will find this all too familiar.

Digital NLE tools were unquestionably faster: changes that once took hours could be accomplished in minutes. And yet, schedules did not collapse the way management expected, because the time that was saved was put to work. Editors used that additional time to try another version, and yet another, delivering output better than originally imagined.

More, or Better

If AI compresses the non-creative hours — the preparation, the organization, the transcription, the cleanup — then that time has to go somewhere. A company can take those hours back, shorten the schedule, and turn around more of what it’s already making at a lower cost.

That time can also be invested. The same hours can go back to the editorial teams, resulting in not just more content but better content.

75.8% of editors said the best use of AI in post production is to give them more time for storytelling and creative decision-making. In a market drowning in volume, the competitive advantage of AI isn’t more content, it’s better content. It’s the thing that makes an audience choose one show over the eleven others queued up behind it.

The Bottom Line

Despite the fact editors have embraced AI for certain tasks, there is still a trust gap, and one respondent described it best:

“How, when using integrated tech, do we trust that the AI companies are not using our daily work to train in order to develop technology that will one day try to replace working post professionals?”

Ethical use, copyright or stolen content, and security concerns were named the number one barrier to adoption by 49.2% of all surveyed — more than triple the next answer, existing tools that don't integrate well (15.0%). Concerns about creative quality, trust in results and a lack of clear benefit each drew 13.3% or less. Editors' wariness is less about what these tools can do than about what data the models were fed, and who profits from it.

Our takeaway is clear: Editors are adopting AI faster than many assume and are spending whatever time they've gained the same way their predecessors did — on making the best story possible. Those studios and productions that can harness the efficiencies behind AI while providing ethical standards and safe adoption won’t just save time and money — they’ll deliver a better final cut.