HumanSignal, from HumanSignal (formerly known as Heartex), is a Chrome extension that flags likely AI-written LinkedIn content. It analyzes feed text and surfaces an at-a-glance assessment of whether a post or comment appears human-authored or generated by an AI model. The tool emphasizes private, fast classification, inline confidence cues, and brief transparency indicators that explain the assessment. It targets recruiters and professionals who want quick authenticity checks while browsing LinkedIn.

What the tool places next to posts and comments

The extension adds small, non-intrusive badges beside individual posts and comments that communicate a classification and confidence level. Labels range from "Feels Human" to "Almost Certainly AI" and are accompanied by transparency indicators that explain the model's confidence. The visual approach aims to give a readable signal without replacing manual judgment, and the badges are designed to occupy minimal screen space within LinkedIn's layout.

How it behaves as you scroll long feeds

HumanSignal performs real-time content analysis while you navigate the LinkedIn feed, using a fast rules engine intended for low-latency classification. That design means assessments appear quickly as new items load, so badges show up during normal browsing rather than after a noticeable delay. The classification path focuses on speed to avoid interrupting the browsing rhythm for users who review many posts.

How it handles privacy and data

The extension uses on-device AI processing so all text analysis runs locally inside the browser, and no post content leaves your machine. The developer states that the architecture prevents external servers from receiving LinkedIn data, and the extension does not store feed text externally. This privacy-first approach suits users who prefer to keep social data confined to their desktop or laptop environment.

Who benefits most and what its limits are

HumanSignal is optimized specifically for LinkedIn and runs on Google Chrome and other Chromium-based browsers across desktop platforms. The developer has a background in the machine learning lifecycle, which shapes the classifier design. Detection is described as a guidance tool rather than definitive proof, so professionals who need a quick authenticity cue benefit most, while users requiring forensic certainty must use additional verification methods.

A practical, privacy-focused cue for professionals who value quick authenticity signals

HumanSignal is a practical option for LinkedIn users who want private, fast cues about whether posts or comments look AI-generated. Its trade-off is that the classifier offers probabilistic assessments rather than conclusive proof, so it works best as a first-pass verification aid. For professionals prioritizing quick, local checks of content origin, it suits routine use; high-stakes decisions require further validation.