LinkedIn adds AI‑slop report button, drops its writing assistant

LinkedIn rolled out a “seems like AI slop” reporting button on its feed. The feature aims to curb the surge of low‑quality AI‑generated posts that clutter professional timelines.

The button appears next to the existing “Report” dropdown and is labeled “Seems like AI slop” (TechCrunch). At the same time, LinkedIn is pulling its internal AI‑writing assistant from the compose pane and swapping it for a lightweight proofreading overlay (Engadget). The changes arrive while the platform reports a steady rise in AI‑crafted content, according to internal data shared with the press.

The new reporting tool

LinkedIn’s reporting option lets users flag a post with a single click. Once selected, the post is routed to a moderation queue that applies a higher‑severity label than a generic report. The label signals that the content likely originated from a generative model and failed basic quality checks.

The rollout is limited to a subset of users for now. LinkedIn says it will expand the button after monitoring false‑positive rates. Early feedback from the beta group indicates that the button reduces the time moderators spend on low‑effort spam, but the company has not released concrete metrics (TechCrunch).

From AI writer to proofreading

LinkedIn’s AI‑writing feature, launched earlier this year, auto‑generated post drafts based on a short prompt. The tool was marketed as a productivity aid for busy professionals. In practice, many drafts sounded generic and required heavy editing.

The new proofreading overlay focuses on grammar, spelling, and style rather than content generation. It highlights errors in real time and offers suggestions without rewriting the entire post. LinkedIn frames the shift as a move toward “quality‑first” assistance, but the change also sidesteps the liability of publishing AI‑generated statements that could misrepresent users (Engadget).

Why AI‑slop matters for the platform

Professional networks rely on signal quality to maintain credibility. When AI‑generated posts flood the feed, they dilute the relevance of genuine expertise. Recruiters, investors, and engineers all scan LinkedIn for authentic insights; a rise in boilerplate content threatens that value proposition.

Moderation at scale is already a pain point for social platforms. Automated detection of AI‑generated text is still error‑prone, and manual review is costly. By giving users a direct way to flag suspect content, LinkedIn hopes to offload part of the burden onto the community. The approach mirrors similar tools on Reddit and Twitter, but LinkedIn’s focus on professional discourse makes the stakes higher.

Industry reaction and next steps

Competitors have taken mixed stances. Some AI‑content platforms argue that labeling AI output as “slop” discourages experimentation. Others applaud LinkedIn for confronting a problem that many tech blogs have warned about since large language models became widely accessible.

Analysts note that LinkedIn’s move could set a de‑facto standard for other niche networks. If the reporting button proves effective, we may see similar controls on GitHub Discussions, Stack Overflow, or even enterprise intranets. The key question is whether the signal from user reports will be enough to train more accurate detection models.

What to watch

LinkedIn plans to publish a post‑mortem on the button’s impact later this quarter. Track the false‑positive rate disclosed in that report and any subsequent tweaks to the moderation pipeline. Also monitor whether LinkedIn expands the proofreading overlay into a full‑featured editor or re‑introduces a toned‑down writing assistant. Those decisions will reveal how the platform balances user convenience with content integrity.