Most wrong-AI uses turned into straight copying instead of correction

Workflow & automation

73.2%

of trials with wrong AI output ended in people following it across the pooled studies

Reader Note:: These shares apply only to trials where participants used the AI and the AI was wrong.
Evidence:: Experimental
Vendor interest:: None
Sample:: Three preregistered experiments with 1,372 total participants in pooled analysis; university lab and Prolific samples solving 7 CRT reasoning items, with participants randomly assigned to AI access and other conditions.

Good for

Backing budget and pricing decisions in similar situations.

Situations far from the one tested, without checking again.
This was measured in one market. It may not hold where you are.
This is an aggregated or average number that covers very different cases. Real results in a particular segment may differ.
The source never says when this data was collected. It could be considerably older than the publication date suggests.

10%

of organisations are realising significant ROI from agentic AI in Deloitte’s 2025 survey of 1,854 EMEA executives

24–56%

of marketing tasks are currently performed using AI or ML, rising to over half in three years according to US CMOs (2026 survey)

Our Substack turns the data behind these cards into full breakdowns: what changed, why, and what to do about it.

90% of CMOs are experimenting with AI but fewer than 10% have scaled it or realized value across marketing workflows

42.3%

of wrong AI answers were overridden with incentives and item-level feedback in Study 3 versus 20.0% in the control group

14%

average increase in sales productivity reported from using AI in marketing by US companies (2026 survey)

77% of brands say AI-assisted influencer campaigns perform better