Substack Launches AI Transparency Tool to Show How Much of Newsletters Are Machine‑Written

Substack has rolled out a new feature that lets subscribers see an estimate of how much of a newsletter was generated with artificial intelligence. By clicking a small icon next to any post, readers receive a percentage range indicating the portion of the content likely produced by AI‑assisted tools. The company says the move aligns with its commitment to “trust and transparency,” giving audiences clearer insight into the origins of the material they consume.

The tool relies on advanced text‑analysis algorithms and machine‑learning models that scan for stylistic markers, repetitive phrasing, and other signatures typical of AI‑generated prose. Results appear as a range—e.g., “30‑40% AI”—signaling that roughly a third of the piece was crafted with the help of automated systems. Substack cautions that these figures are approximations, not definitive measurements.

The launch comes amid growing scrutiny over the rise of AI‑aided writing across blogs, newsletters, and news outlets. Readers, especially in sectors like finance, health, and technology, are demanding more clarity about whether the information they receive is human‑authored or machine‑assisted. Substack’s new indicator not only pushes writers to disclose AI usage but also equips editors with a built‑in compliance check for ethical publishing standards.

Initially, the feature is available to a limited group of beta testers, with plans to expand platform‑wide over the next few months. Substack acknowledges that AI‑generated content is becoming a standard part of the publishing workflow, but insists that transparency is essential for maintaining subscriber trust. The development may spark further debate about the role of artificial intelligence in digital media and the responsibilities of platforms to keep readers informed.

Source: TechCrunch

ETİKETLER