All figures in this article come from our 2026 video study, based on responses from 688 marketing and communications professionals.
87.4% of marketing and communications teams already use AI in their video workflow. On paper, adoption looks nearly universal.
In practice, AI is still mostly used for supporting tasks. It writes subtitles, suggests script ideas, or translates a voice-over. It rarely creates the video itself. And time saved on individual tasks does not always lead to faster publishing.
What Teams Use AI For
Teams primarily use AI for subtitles, translations, and voice-overs (58%), followed by idea generation and video scripts (57%). Creating or editing the video itself trails well behind at 36%.
Content repurposing (clips, summaries) comes next at 30%. 12% of teams do not use AI at all, and 10% have tried it without success.
The ranking draws a clear line: today, AI handles what surrounds video production (text, translation, ideation) more than the video itself.
Does AI Really Save Time in Video Production
75.6% of teams say AI saves them time in video creation.
But that figure hides a more nuanced reality:
- only 30.8% describe those time savings as significant (up to cutting production time in half),
- around 45% say the gains are modest,
- and 14% see no real impact.
So yes, AI helps most teams. But it rarely changes the game on its own. Part of the issue is measurement: 42.2% of teams have no clear video KPIs, and only 28.2% say they can connect video performance to business results.
Why Time Saved Does Not Automatically Mean Faster Publishing
The time AI saves often gets lost elsewhere in the workflow: approval cycles, handoffs between teams, and switching between tools.
AI saves time task by task, but those gains do not automatically add up to faster publishing.
The real bottleneck is not how fast one isolated task gets done. It is how smoothly tasks connect across the full workflow. And that flow has a direct impact on performance: teams that finish a video in under a day are 5 times more likely to see very strong performance than teams that take more than a week.
Why So Few Teams Use AI to Create the Video Itself
Only 36% of teams use AI directly in the creation or editing stage, compared with 58% for subtitles and 57% for ideation.
That gap matters. It shows that trust in AI is still much stronger for support tasks that are structured, guided, and easy to review than for the act of creation itself.
As long as that trust ceiling stays in place, widespread AI adoption will continue to coexist with production workflows that remain, for the most part, unchanged.
This trust ceiling is not unique to AI. It reflects how teams structure creation more broadly: 63.7% rely mostly or entirely on specialists, and only 9.9% let anyone create autonomously.
