59% prefer the brand that uses AI. 74% want to know when it does

Using AI doesn’t drive the Brazilian consumer away, but they want to know when it happens. In the Globo survey from August 2026, 59% prefer a brand that uses AI over a similar one that doesn’t. At the same time, 74% want to be notified when an ad uses AI, and 95% defend that all AI-generated content be labeled.

What Globo measured

The survey The impact of AI on Brazilian consumption habits was conducted by Globo with E&I Inteligência Setorial, with 1,000 interviews in August 2026. For anyone who runs communication, the contrast is between preference (59% choose the brand that uses AI) and the request for notice (74% want to be informed in ads). Labeling AI-generated content brings together almost everyone: 95%.

Preferring the brand that uses AI and wanting to know when it does coexist without contradiction. The consumer accepts the technology and demands transparency.

What the data doesn’t say

The survey doesn’t measure what happens to a brand caught using AI without notice. Any number on that would be a guess. What it shows is the request: almost everyone wants labeling. In our reading, giving notice costs a line of text, and the cost of being found out is an unknown that isn’t worth testing.

What the platforms already do

Google advises that, when it makes sense for the audience, the site explain how the content was created, including the role of automation. It also asks for manual review of everything AI generates before publishing. LinkedIn, since 2024, labels AI-generated images with a badge in the corner, automatically, through the C2PA standard. The article we consulted doesn’t mention a label for text.

The direction is the same the consumer asked for: signal it.

A simple AI-use policy

This is our proposal, not a rule. Four rules cover most cases:

1. AI-generated image, video and voice: always signal.
2. Text written by AI and published without relevant editing: signal.
3. Text with an AI draft, reviewed and signed by a person: sign with the name of whoever is accountable for the content.
4. Analysis done with AI over data: cite the source of the data, not the tool.

Also define who approves and where the policy is published. It only holds if the team knows it and the customer can read it.

Examples of signaling

• “Image created with AI.”
• “Draft made with AI, reviewed and signed by [name].”
• “Analysis done with AI over [source, year] data.”
• “Voice synthesized by AI.”

Short lines, at the end of the ad or post, or in the image caption. No apology and no alarm tone.

To see how AI communication fits the marketing routine, read about content with AI without losing authenticity and the CMO’s role in the AI era. NorthStar sets up that flow on the marketing intelligence page.

Frequently asked

Do I need to disclose that I used AI in an ad?

There’s no single rule for every case. But 74% of Globo respondents (Aug/2026) want to be notified when an ad uses AI, and 95% defend labeling AI-generated content. Disclosing is the choice aligned with what the public asks for.

Does using AI hurt the brand's image?

In the Globo survey, 59% prefer a brand that uses AI over a similar one that doesn’t. The figure measures preference, and doesn’t measure the reaction to those who hide the use.

Does Google penalize AI-written text?

Google accepts AI-made content. What it treats as spam is producing many pages with no value to the user. It recommends manual review and, when it makes sense, explaining how the content was created.

Does LinkedIn label AI content?

It labels AI-generated images, through the C2PA standard, since 2024. The article we consulted doesn’t mention a label for text.

Sources

• Globo Gente / E&I Inteligência Setorial. The impact of AI on Brazilian consumption habits, 09/10/2026. gente.globo.com
• Google Search Central. Using generative AI content on your website. developers.google.com
• Social Media Today. LinkedIn adds labels for AI-generated content, 2024. socialmediatoday.com