Skip to main content
← Back to market wire
FundingTechCrunch

LinkedIn adds a button to report AI-generated ‘slop’

LinkedIn is introducing new ways to reduce low-quality AI-generated posts, including a “seems like AI slop” reporting option. It's also replacing its own AI writing feature with a proofreading tool.

Desk analysis

AI-assisted2 min read

LinkedIn has decided that the most efficient way to police its own feed is to ask the audience to do it. A new "seems like AI slop" reporting button now sits alongside the familiar flags for misinformation and harassment, giving users a one-click way to flag the synthetic prose that has quietly colonized the platform.

The move is a confession dressed as a feature. LinkedIn spent two years encouraging members to "rewrite with AI," and the result was a feed thick with the same vaguely authoritative tone, the same hollow confidence, the same bullet-pointed wisdom that nobody actually believes. Now the company is replacing its AI writing assistant with a proofreading tool, a quieter utility that fixes grammar without manufacturing thought.

The economics are straightforward. Engagement metrics reward volume, and AI generates volume cheaply. The cost is trust, which is harder to measure and slower to recover. By outsourcing moderation to users, LinkedIn buys time without committing to the harder work of downranking machine-made content at the source.

For the broader labor conversation, the signal is worth noting. LinkedIn remains the default résumé and recruiting surface for white-collar work, and the platform's content shapes how professionals present themselves. When the venue itself cannot distinguish a human voice from a generated one, recruiters and candidates alike are reading through a haze. A reporting button is a start; it is not a solution.

The interesting question is what comes next. If enough users flag the slop, LinkedIn will have data on which accounts and which patterns produce it. That dataset is more valuable than any moderation policy, because it points directly at the automation layer the platform will eventually need to throttle. The button is the user-visible tip; the underlying machinery is the real product.