A recent analysis of popular Substack newsletters revealed that around 10% of them utilize AI-generated or AI-assisted content. This finding comes as a surprise given Substack’s subscription-driven model, which differs from other platforms like Facebook and YouTube that prioritize engagement and traffic. The AI-detection startup GPTZero surveyed 25 to 30 recent posts from the top 100 Substack newsletters, confirming that 10 of the newsletters incorporate AI, with seven relying heavily on it.
Most of the flagged newsletters provide financial advice and investment news, indicating that a significant audience is consuming AI content without realizing it. Alex Cui, CTO of GPTZero, noted the unexpected prevalence of AI use among prominent authors on Substack, pointing out that the rate is higher than similar analyses conducted on Wikipedia.
Not all observers are surprised, however; some believe that the integration of AI in financial newsletters makes sense when considering the nature of the content they produce. Substack’s head of communications, Helen Tobin, acknowledged the platform’s mechanisms for identifying spam but clarified that they do not actively remove content based solely on its AI origins, recognizing legitimate uses for AI in content creation.
Despite the growing presence of AI content, Substack appears to have a lower rate of such writing compared to other platforms. For instance, previous analyses found that about 40% of content on Medium utilized AI. However, while much of that content garnered low engagement, the AI-assisted writing on Substack comes from reputable sources.
As Substack positions itself as an alternative to mainstream media, the increase in AI-generated writing reflects a broader trend observed across various news outlets. Notably, some AI tools have already been adopted by well-known publications for tasks such as article summaries.
While many readers show indifference to the use of AI in their favorite newsletters, concerns about authenticity are surfacing. In response, GPTZero plans to introduce a “certified human” badge for blogs, indicating content that is free from AI assistance—an initiative that mirrors disclaimers appearing in other creative fields, emphasizing the growing dichotomy of human versus AI content.
This evolving landscape suggests that while AI’s presence in media continues to grow, the conversation on ethics and transparency in content creation is likely to intensify in the coming years.
