AI adoption in content creation isn’t uniform across formats, and the gaps are wider than you might expect. A survey of 100 companies found AI deeply embedded in image production, with adoption in text and audio close behind. Video is the clear outlier, trailing the average by nearly twenty percentage points. That gap suggests video remains the format where human production still holds the most ground, likely due to complexity and quality expectations.

Essential Statistics

• AI integration in image creation reaches 91.4%, the highest of any format measured.
• AI integration in text creation reaches 84.1%.
• AI integration in audio creation reaches 82.8%.
• AI integration in video creation reaches only 59.7%, the lowest of any format.
• The average AI integration level across all four formats is 79.5%.

Key Takeaways

• Image creation leading in AI integration suggests visual content generation has matured faster and become more reliable than other formats.
• Text and audio both sitting close to the overall average suggests these formats have reached a similar level of AI maturity and comfort among marketing teams.
• Video trailing well behind every other format suggests either technical limitations or quality concerns are still holding back full AI adoption in that medium.
• Since video adoption sits nearly twenty points below the average, this format likely still requires more human oversight than image, text, or audio production.
• This gap is worth watching, since video AI tools are advancing quickly, and integration levels here could shift meaningfully within a relatively short window.

Actionable Insights

• If your team hasn’t integrated AI into image production yet, this is the format where adoption is furthest along across surveyed companies, and it’s a reasonable low-risk starting point.
• Approach AI video adoption more cautiously than other formats, given how far behind it trails the average. Testing on lower-stakes video content before scaling is worth doing rather than assuming the same workflow that works for text will translate directly.
• Benchmark your own AI integration levels across image, text, audio, and video against these figures to identify which format represents your biggest opportunity or biggest gap relative to peers.
• Keep closer human oversight on video production specifically, since the lower adoption rate here suggests the format still requires more manual quality control than others.
• Revisit your video AI tooling every few months rather than assuming today’s limitations are permanent. Given how quickly other formats caught up, video adoption is worth reassessing periodically.

Video is lagging for a reason. It’s the format where AI still can’t fake the details people notice, and until it can, human hands are staying on it. – Neil Patel