LinkedIn adds reporting for AI-generated content and removes its AI writing feature

On July 30, 2026, LinkedIn announced a set of coordinated changes to how it handles AI-generated content. The most visible is a new user-facing report option labeled “seems like AI slop,” letting members flag posts they believe were generated by an AI tool rather than written directly. Behind the scenes, LinkedIn is deploying detection classifiers to reduce the weight given to AI-generated content in the recommendation algorithm and adding automated defenses against bot-generated comments.

Alongside these changes, LinkedIn is removing its “enhance your post” AI writing feature — which produced full post drafts from a short prompt — and replacing it with a proofreading tool. The distinction the company is drawing is between AI as a generator of content, which it is removing, and AI as an editor of content, which it is keeping.

For product managers working on social, community, or communication features, this is a notable directional signal. The enforcement move suggests that enough users have become sensitive to AI-generated content to complain about it at meaningful scale — and that LinkedIn believes addressing this is worth removing a feature the platform itself had shipped. Crowdsourcing detection through a report button makes the system self-reinforcing: user flags generate labeled data, which improves the classifier.

The distinction between generation and refinement is becoming a practical design question across multiple platforms. Substack and other writing platforms have been navigating similar dynamics in 2026. Teams building writing assistance features may need to revisit whether generation capabilities create trust problems in contexts where community authenticity matters to the audience, even if individual users are willing to use the tool.