AI transparency in newsletters: what changes for content marketing

Updated October 16, 2026

83% of marketing executives already use AI in the creative process, but only 45% of consumers have a positive perception of AI-generated content. Substack has just made this gap visible to everyone.

Does this news really affect me, or is it just an issue for American creators?

It concerns anyone who uses a newsletter as a channel to engage clients and prospects, in Italy and elsewhere. When a platform with millions of creators adopts a standard, others follow.

On July 22, 2026, Substack made a new feature public: a tool to estimate the extent to which a piece of content was written or assisted by artificial intelligence. As reported by TechCrunch, the platform has integrated Pangram, an AI detection software that analyzes posts, notes, replies and comments of at least 100 characters and returns a percentage estimate of AI-generated content.

The result is visible only to the user who requests the scan, not to the general public. But the signal is clear: AI transparency is no longer optional.

For an Italian marketing manager, this is not a technical detail. It directly affects your brand's credibility and your relationship with the people who read your communications.

How exactly does Substack's Pangram tool work?

Pangram analyzes texts published on Substack after July 2026 and estimates the share of AI-generated content. Creators can add a voluntary label or dispute an incorrect classification.

Pangram is AI detection software, meaning a program that analyzes the style and structure of a text to estimate whether it was written by a human or generated by a language model such as ChatGPT or Claude.

According to Gazetteer SF, the feature works only on content published after the July 2026 announcement. Earlier posts remain unverifiable on the platform.

One important point: as MyStartupNews notes, creators can dispute an incorrect classification. This matters because AI detectors aren't infallible: a very polished, formal text can be classified as AI-generated even if a human wrote it.

What the data says: how big is the gap between those who use AI and those who trust AI?

83% of marketing executives already use AI in the creative process, but only 45% of consumers have a positive perception of AI content. That gap is the problem to solve.

The most telling figure isn't 83%. It's the gap between 82% and 45%: advertisers believe almost all consumers feel positive about AI in content, but reality is nearly the opposite. This misperception is costly: it leads to communication choices that erode trust instead of building it.

Trust, after all, is already under pressure. According to the Edelman Trust Barometer 2026, which surveyed almost 34,000 people in 28 countries, nearly 7 in 10 respondents fear that institutional leaders deliberately mislead the public. In this climate, transparency about AI isn't an optional ethical gesture: it's a concrete way to stand out.

Why it's being talked about now: is Substack really relevant for the Italian market?

With over 100,000 monetizing publishers and an average open rate of 44%, double the email average, Substack has become a benchmark for all newsletter content marketing.

In June 2026, Substack passed 100,000 publishers earning through subscriptions, with 5 million active paid subscriptions. The average email open rate is 44%, about double the industry average.

These numbers explain why Substack's move matters even if you don't use Substack. When a platform of this scale adopts a standard, in this case AI disclosure, it effectively turns it into a market expectation. Beehiiv, LinkedIn Newsletters, traditional email marketing platforms: they will all have to answer the same question.

For an Italian company that uses its newsletter to build relationships with customers and prospects, the question isn't “Should I be on Substack?” The question is: “Is my AI content policy clear, consistent and communicated?”

This applies to anyone running paid media campaigns too. As we discussed in our piece on Gemini and AI image generation for content marketing, AI is already part of almost everyone's creative process. The real question is how to communicate it.

What the law requires: is the EU AI Act already in force?

Yes. As of August 2, 2026, Article 50 of the EU AI Act applies and imposes disclosure obligations for AI-generated text on matters of public interest, with fines of up to 15 million euros.

The AI Act, the European regulation on artificial intelligence, is no longer a future prospect. As Baker Botts explains, Article 50 became applicable on August 2, 2026 and covers anyone who publishes AI-generated text on matters of public interest.

“Public interest” is a broad category: it includes health, finance, law and politics, but also communications that influence major purchasing decisions. If your company produces content on these topics with AI support, disclosure is no longer a choice.

The rule doesn't ban the use of AI: it requires you to disclose it. For many Italian companies, the first step is to work out which content falls within its scope and update internal policies accordingly.

Three real-world scenarios: what changes for an Italian SME that uses newsletters and AI content?

The impact depends on how much AI you already use and how you communicate with your audience. Here are three typical situations with recommended actions.

Illustrative examples: the numbers and situations are meant to show the reasoning, they are not real customer data.

What tools and approaches exist to manage AI transparency in content?

There are three main approaches: voluntary disclosure, automatic detectors like Pangram, and internal editorial policies. The best solution often combines them.

The right choice depends on your content volume, industry and audience. For a chain of stores sending promotional newsletters, a standard label and an internal policy are enough. For a law firm publishing regulatory analysis, certified human review is almost a must, including for AI Act compliance.

It's also worth looking at how other channels are evolving: the issue of digital content credibility also touches on the fight against content theft that X and Grok are facing, and more broadly on how platforms are redefining the rules of the game for creators.

What real risks do you run if you ignore this trend?

There are three risks: regulatory (AI Act), reputational (loss of trust) and competitive (those who are transparent stand out). Ignoring them won't make them go away.

Regulatory risk is the most immediate: Article 50 of the AI Act already applies. But reputational risk has the longest-lasting effects.

The Edelman Trust Barometer 2026 shows that nearly 7 in 10 people fear being deceived by institutional leaders. In this climate, a brand caught using AI in its content without disclosing it, even unintentionally, can suffer credibility damage that is hard to recover from.

Where to start: practical steps to manage AI transparency in your communications

Start with an internal audit: understand where and how you already use AI in your content, then define a policy and communicate it. Five steps, in order.

What mistakes are made most often when tackling this topic?

The most common mistake is waiting for the problem to show up instead of defining a policy first. The second is treating disclosure as a bureaucratic obligation rather than a brand lever.

Many Italian companies are already using AI in their content, often without an explicit policy. This creates three typical problems:

• Internal inconsistency: every team member uses AI differently, with no shared rules on review and disclosure.
• Reacting instead of planning: you act only when a complaint or a wrong classification arrives, instead of preventing it.
• Disclosure as an excuse: you add an AI label without caring about content quality, losing the competitive advantage of transparency.

A less obvious mistake: believing disclosure hurts the brand. IAB data shows that 72% of marketers want an industry standard, which means transparency is becoming the expected norm, not the exception.

Video and multimedia content are affected too: as we saw when analyzing YouTube's changes to view counting, platforms are redefining credibility metrics across all formats.

How can you tell whether your AI policy is really working?

Three indicators to monitor: newsletter open rate, qualitative reader feedback, and the absence of misclassifications by automated detectors.

So, is AI transparency really worth it for your company?

It's worth it if you already use AI in your content (even just for drafts or editing), publish newsletters or content on regulated topics, or want to stand out from competitors with more authentic communication.

A middle path: you use AI only for internal (unpublished) work and don't operate in high-regulatory-risk sectors. You can wait a few months, but prepare your policy now.

Not yet if you produce all your content entirely without AI and have no plans to introduce it. Still, keep an eye on how standards evolve, because the market moves fast.

How we see it at SHM

For us, AI transparency isn't a compliance issue: it's about respect for the people who read. We believe AI should be integrated to free companies from routine, not to replace human ingenuity or to produce content nobody has really thought through.

When we work with companies on strategic AI integration, the first step is always defining where AI adds real value and where it risks eroding the relationship with your audience. Disclosure isn't a cost: it's the signature of someone who still has something to say.

If you want to understand how to structure an AI policy for your content, or how to position your brand in the new ecosystem of Marketing and performance ADV, we're available for a concrete conversation. You can also learn how to optimize your presence for AI search engines with our SEO for AI service, or train your team with our AI training for companies program.

Content credibility also touches visual marketing: if you're considering how to use AI to produce images, read our analysis of Adobe AI Critique and visual marketing.

Frequently Asked Questions

AI transparency isn't the problem: it's the answer to the trust problem that was already eroding before.