Four in 10 consumers say their trust in a brand drops when they suspect a marketing email was written by AI, according to a report cited by Digital Today.

The 40 percent figure lands on an industry that has moved AI email generation from pilot into production. Every major email service provider now ships some form of generative copy tool. The consumer backlash number suggests automation rollout is moving faster than buyer tolerance.

How should operators read the 40 percent?

Digital Today's headline reports trust erosion, not full rejection. Three interpretations matter for email program owners:

• The number counts consumers who have noticed AI copy. Detection rates keep climbing as buyers become familiar with the typical patterns of machine output.
• Trust loss converts to measurable behavior — lower click-through, faster spam-folder routing, and quicker unsubscribe.
• The remaining 60 percent span a wide range. Some buyers cannot tell the difference. Some segments appear more tolerant of machine-drafted text.

The single statistic does not tell operators where their own audience sits. Treat it as a directional warning, not a precise target.

What drives the trust drop?

Three mechanisms appear consistent with operator experience, even where the Digital Today report does not break them out:

• Homogenization. Generative copy converges on similar sentence structures, vocabulary, and calls to action across brands. Buyers notice when every company sounds the same.
• Disclosure expectations. Hidden machine authorship reads as deception. Consumers who want AI flagged react strongly when they find it unflagged.
• Lost brand voice. Voice is a multi-year investment. Machine copy flattens it within minutes.

For a channel where revenue per send runs in cents and list-churn compounds, these mechanisms carry financial weight.

What should email teams change?

The 40 percent number carries four operational implications for teams running email programs:

• Classify your sends. Pull a 100-message sample and tag each as human-written, AI-assisted, or AI-generated. The trust risk sits mainly in the third bucket.
• Test disclosure. Split a campaign into AI-labeled and unlabeled variants. Measure open, click, and unsubscribe deltas over 30 to 60 days.
• Reserve AI for structure, not voice. Use machine output for subject-line variants, send-time picks, and segmentation logic. Keep body copy under human authorship.
• Track list health weekly. Watch spam-complaint rates and unsubscribe velocity. A spike often signals that AI copy has tipped past the trust ceiling.

The split between assistive use and generative use is not cosmetic. Buyers who accept AI as a drafting tool still reject it as a substitute for a human editor.

What the headline does not tell us

Digital Today's report cites the 40 percent figure without breaking out:

• Survey sample size
• Methodology
• Industry or geographic split
• Whether respondents knew the emails were AI-written or guessed

Operators should treat the number as a directional signal, not a precise benchmark. The headline alone justifies pausing fully automated email copy until segmentation data surfaces.

Bottom line for operators

Email is the workhorse direct channel for most retailers and B2B software vendors. A trust drop that pushes 40 percent of buyers toward spam-folder routing or faster unsubscribe will erode list ROI quickly. Teams that ship AI copy without disclosure are running an uncontrolled A/B test with their own list health as the dependent variable.