New research from Optimizely, surveying over 2,000 marketing leaders across seven countries, surfaces a problem a lot of teams are feeling but haven't named yet: AI is producing more content, faster — but it isn't actually saving anyone time.
The Revision Tax
The work didn't disappear. It moved — from creation to correction. Call it the revision tax: the hidden cost of cleaning up after AI that eats into the efficiency gains it was supposed to deliver.
The Pressure to Ship Anyway
That correction burden creates a second problem: pressure to publish under deadline, imperfections and all.
- 25% admit they've published off-brand AI content because of time constraints
- 30% have passed off AI-generated work as their own
- Only 19% work from a single, integrated AI platform — most teams are stitching together disconnected tools, which only adds friction
The Leadership Blind Spot
Maybe the most interesting finding is the gap between how leadership thinks AI adoption is going and how it's actually landing with the people doing the work.
69% of C-suite leaders believe AI adoption is well-aligned across their organization. Only 27% of analysts — the ones actually in the tools every day — agree.
There's an irony here too: C-suite leaders are the most likely to pass off AI work as their own (44%), nearly double the rate of managers.
Is AI Costing Marketers Their Voice?
- Only 30% feel their brand voice still feels genuinely distinctive
- 46% worry AI is eroding the creative skills of junior marketers
- 53% say AI can capture the facts of a brand but not its emotional resonance
The Takeaway
AI adoption isn't failing — it's just further along on the "hype vs. reality" curve than most orgs are willing to admit. The tools are generating volume. What they're not doing is replacing judgment, brand instinct, or the review work that keeps output on-brand and factually sound.
The fix isn't more AI. It's fewer disconnected tools, tighter feedback loops between the people using AI and the people setting strategy, and being honest about where the "efficiency gains" are actually going.