AI in web content: what the Pew study means for SMEs

The context: half a million pages analyzed

At the end of August 2026, the Pew Research Center published a wide-ranging study. Researchers examined nearly 500,000 English-language web pages. The goal was to quantify how widespread AI-generated text is on the commercial web. The results, reported by The Decoder, confirm clear and measurable growth.

The study focuses on the period after ChatGPT launched in late 2022. Since then, AI-assisted content production has accelerated at an unprecedented pace. The Pew data won't surprise industry insiders, but for the first time it provides a solid quantitative baseline.

As with other technology trends, adoption is uneven. Commercial .com sites show ten times the concentration of AI content found on institutional domains (.edu, .gov). This matters for anyone working in digital marketing and SEO.

The numbers that really matter

The headline finding is clear: over a third of pages published after November 2022 show signs of machine-generated text. Still, it's important to put that figure in context. Not all AI text is equal in quality or editorial value.

The concentration on .com domains also suggests that commercial pressure is the main driver of adoption. Companies want more content at lower cost. As a result, many sites have built automatic generation tools into their editorial workflows.

The Pew study captures only the beginning of a structural shift in how companies produce digital content.

Finally, it's worth noting that the sample is in English. The Italian market is still slightly behind the English-speaking one, but the adoption curve is similar. Italian SMEs still have room to position themselves strategically.

The strategic view: what changes for SEO and content marketing

The spread of AI content has a direct effect on organic rankings. Google has said repeatedly that it doesn't penalize AI-generated content as such. It penalizes low-quality content, whatever its origin. In practice, though, the line is thin.

Google's evaluation systems, summarized in the E-E-A-T framework (Experience, Expertise, Authoritativeness, Trustworthiness), favor content that shows firsthand experience and authority. These are hard to replicate automatically. An unoptimized AI text risks coming across as generic and adding no value.

At SHM Studio, we see this every day in the SEO projects we manage. Sites that adopted AI content indiscriminately often see organic traffic drop in the following months. By contrast, those who combine AI with structured editorial oversight get stable or growing results.

So the critical variable isn't the use of AI, but the quality of the overall editorial process. That includes defining a digital marketing plan aligned with your business goals.

The commodity risk: when all content looks the same

The Pew study points to a side effect that deserves specific attention. If over a third of the commercial web uses AI text, the risk of content homogenization becomes real. Similar articles, identical narrative structures, standardized vocabulary.

Language models also tend to converge on probabilistically central answers. As a result, texts generated by different tools, such as GPT-4o, Claude and Gemini, often show overlapping stylistic patterns. For a brand that wants to stand out, that's a strategic problem.

The competitive value of digital content is shifting toward authenticity and a proprietary point of view. Companies that invest in a distinctive editorial voice will gain a growing edge over those relying solely on automation.

In this sense, strategic copywriting isn't a cost to cut, but a competitive asset. The difference between copy that converts and copy that gets ignored often lies in details only an experienced writer can catch.

Operational implications for Italian SMEs

Turning the Pew data into concrete decisions takes a shift in perspective. It's not about deciding whether or not to use AI. It's about defining where AI adds value and where it takes it away.

Here are some practical guidelines for marketing managers:

• High-volume, low-complexity content such as product pages, FAQs and category descriptions are natural candidates for AI-assisted automation. They still need editorial review before publication.
• Strategic content such as case studies, whitepapers and thought leadership articles must keep a dominant human component. These formats build authority and can't be handed entirely to AI.
• Blogs and content hubs call for a hybrid strategy. AI can support research, structure and first drafts. Editorial review ensures quality and consistency with the brand's tone of voice.
• Paid campaigns on Google Ads or LinkedIn benefit from AI for creative testing, but need strategic oversight to avoid generic messaging.

It's also essential to monitor content performance over time. A text that performs well today could lose rankings if the search engine updates its evaluation criteria. That's why SEO should be seen as an ongoing process, not a one-off intervention.

A Milan agency's perspective: what the data doesn't say

The Pew Research numbers help us grasp the scale of the phenomenon, but they don't tell the whole story. They don't distinguish between high-quality and mediocre AI content. They don't measure the impact on user experience or user trust.

In our direct experience with Italian SME and mid-market clients, the problem isn't AI itself. It's adopting it without a strategy. Many companies started using automatic generation tools without setting editorial guidelines, without training their teams and without aligning the new workflows with their existing digital marketing goals.

Likewise, not every agency offering AI applied to marketing has the same methodological expertise. The difference shows in the results at six and twelve months. Choosing the partner to navigate this transition is itself a strategic decision.

Finally, consider reputation. When a third of the commercial web uses AI text, brands that communicate with authenticity and depth stand out by contrast. Italian SMEs can seize this competitive advantage today, before the market consolidates further.

Looking ahead: what to expect in 2027-2028

The trajectory is clear. The gap between companies that have built a structured approach to AI in their communications and those still in the exploratory phase tends to widen over time.

Over 2027-2028, it's reasonable to expect search engines to get more sophisticated at detecting AI content. As a result, editorial quality will matter even more for organic rankings. Users themselves will also grow more sensitive to content that feels authentic.

For Italian SMEs, the time to build a hybrid, sustainable content strategy is now. Waiting means chasing a market that moves fast. The SHM Studio team is ready to support you through this transition with a consulting approach grounded in data and hands-on experience. To learn more, check out our blog or get in touch directly from the contact page.