Distrust doubled in twelve months. Then the labels became law.

The pitch was irresistible and nearly universal: generative tools would collapse the cost of content to near zero, so publish more of it — more product pages, more social posts, more email variants, more everything. Marketing organizations restructured around throughput. Agencies repriced. Two years on, the market has returned a verdict, and it is not the one the throughput thesis predicted. Audiences did not reward the flood. They learned to smell it, and they started charging brands for it.

The clearest read comes from a Q2 2026 Fractl study of 1,008 U.S. consumers and 150 marketers. In 2025, 20% of consumers said heavy AI use would reduce their trust in a favorite brand. In 2026, that figure is 40% — distrust roughly doubled in twelve months, with only 14% saying heavy AI use would make them trust a brand more. Among Gen Z, the cohort every consumer brand is chasing, 54% say their trust would drop. The same study found the perceived helpfulness of AI over traditional search fell from 82% to 54% year over year. Usage kept climbing. Confidence did not.

The ReversalAudiences aren't rejecting AI. They're rejecting sloppy AI.

It would be easy — and wrong — to read this as a consumer revolt against artificial intelligence. DoubleVerify's global survey of 22,000 consumers across 22 markets and more than 2,000 marketers found the opposite at the top line: 63% of consumers globally say AI-powered tools such as search summaries, chatbots and recommendations have improved their online experience. The sentiment is regional and nuanced — 69% positive in LATAM, 68% in APAC, 53% in EMEA, and a notably more skeptical 50% in North America — but the direction is not hostile.

What audiences punish is quality and context. In the same study, 42% of consumers said low-quality AI advertising would negatively affect their opinion of a brand, against just 24% who would feel positively. In the UK the penalty is steeper: 48% would think less of a brand whose advertising ran beside poor-quality, spam-like AI content, and 30% called the impact very negative. More than half of marketers surveyed said they were already worried about exactly this adjacency. Fewer than half of UK consumers say they can reliably identify AI-generated content at all — which is precisely why brands, not algorithms, absorb the suspicion.

The DeadlineOn August 2, disclosure stopped being a brand choice

Sentiment shifts are slow-moving risks. Regulation is not. The transparency obligations in Article 50 of the EU AI Act became applicable on 2 August 2026 — eight days before this brief. Providers of generative systems must mark synthetic audio, image, video and text in a machine-readable format so it can be detected as artificially generated; deployers must disclose deepfakes and certain AI-generated publications. The obligations attach to systems placed on the EU market irrespective of where the provider sits, which means a U.S. brand running EU campaigns is inside the perimeter.

For marketing organizations the practical reading is blunt. A marketing function generating synthetic content for EU campaigns engages Article 50(2); depicting real people — a founder, an athlete, a customer — engages the deepfake labeling duty in Article 50(4). Counsel reviewing the European Commission's final guidance has flagged that the "evidently creative" carve-out brands hoped to lean on is very narrow in advertising contexts. The Commission's Code of Practice on Transparency of AI-generated Content is currently the only EU-wide instrument assessed as adequate for demonstrating compliance, with enforcement running through national market surveillance authorities.

The same standard, three different exposures

The PlaybookPublish less. Prove more.

The brands that will come through this repricing intact are not the ones abandoning AI — they're the ones who stopped treating output as the metric. Four moves define them:

• Label deliberately, not defensively. With 78% of consumers rating explicit labeling as a top trust factor — and demand running at 84% for written content, 91% for video and 90% for images — disclosure is a positioning asset, not a compliance tax. The brands that label first will look confident; the ones dragged into it will look caught.
• Watch your adjacency in real time. You can be penalized for content you never made. Pulse monitors earned, owned and social environments continuously, so you learn where your brand is appearing beside low-quality synthetic content before your audience tells you.
• Ground every asset in live intelligence. Slop is what happens when generation is disconnected from evidence. Spark produces from the same real-time signal your strategy runs on, so speed and substance stop trading against each other.
• Buy verified human authority, not reach. Gartner expects brands to route half their influencer budget into content and creator authenticity work by 2027 — identity verification, provenance checks, deepfake safeguards. Trace identifies the developers, analysts and micro-influencers who genuinely move opinion in your category, and confirms they're real before you activate them.

Bottom LineProvenance is the new brand safety

For a decade, brand safety meant keeping your ad away from the wrong video. In 2026 it means something harder: proving that what you publish came from somewhere, that a human is accountable for it, and that the environment surrounding it hasn't quietly degraded your credibility. Distrust doubled in a year while adoption kept climbing — that gap is the opportunity, because it means the penalty is aimed at carelessness, not at technology. The regulation arriving now simply makes the audience's instinct enforceable.

The strategic consequence is a reversal of the last two years of advice. Content volume has been commoditized to zero and, at the margin, has turned negative. What has appreciated is provenance: evidence, disclosure, verified human authorship, and a trustworthy neighborhood. The brands that win know something others don't — and in this cycle, what they know is that their audience stopped asking whether the content was made by AI and started asking whether anyone was standing behind it.