LinkedIn adds a button to report posts that "seem like AI slop"

LinkedIn Starts Letting Users Report Posts That “Seem Like AI Slop”

LinkedIn is rolling out a new reporting option that gives users a direct way to flag posts they believe were generated by artificial intelligence and feel low quality or inauthentic. The company has placed the option inside the three-dot menu on posts, where users can now mark content as something that “Seems like AI slop.”

The change is notable for two reasons. First, it adds a specific user-facing mechanism for a problem that has become increasingly visible across professional social platforms: generic, formulaic, AI-assisted posting that can overwhelm feeds. Second, LinkedIn is not treating the issue as a small moderation tweak. According to the company’s chief product officer Hari Srinivasan, reducing “AI slop” is a major priority, and the new reporting flow is part of a broader effort to adjust how the platform detects, ranks, and presents content.

For a network built around professional identity, reputation, and industry conversation, the quality of posts matters more than simple engagement volume. LinkedIn’s problem is not merely that some users are experimenting with AI writing tools. It is that a large amount of content can start to feel interchangeable, mechanically phrased, and detached from genuine experience. When that happens at scale, the platform risks weakening the trust that makes professional networking useful in the first place.

A User Signal for Inauthentic Content

The newly added report button creates a feedback loop between users and LinkedIn’s ranking systems. In the reported tests described in the source material, clicking the option hides the post and shows a thank-you message, indicating that LinkedIn is gathering explicit feedback rather than relying only on automated detection.

Srinivasan said the signal will help the company “tune our models and make better feeds.” That suggests LinkedIn is using the feature not just to remove or downrank individual posts, but to improve its broader understanding of what users perceive as inauthentic or low value.

This matters because content-quality problems are difficult to solve with a single rule. A post written with some AI assistance is not necessarily misleading or useless, while a fully human-written post can still be spammy. LinkedIn appears to be aiming for a more behavior-based approach: combining user feedback with classifiers that look for patterns associated with low-quality content.

Why LinkedIn Is Moving Now

The rollout comes amid growing evidence that AI-generated writing has become common on LinkedIn. The source text cites findings from AI detector Pangram, reported by 404 Media, that 41 percent of long-form LinkedIn posts were flagged as being completely generated by AI. Even with the usual caveats around AI detection tools, that figure helps explain why the company is moving more aggressively.

The platform’s challenge is larger than a few awkward posts. LinkedIn has become a place where work announcements, hiring narratives, management advice, and personal-brand storytelling are all heavily incentivized. Generative AI fits that environment almost too well: it can quickly produce polished, confident-sounding text in formats the platform already rewards. The result is a feed that may remain active and highly produced while feeling less personal and less trustworthy.

That erosion of authenticity has become an unusually visible issue because LinkedIn’s identity is tied to real names, career histories, and professional standing. Users generally expect posts to reflect actual perspective, experience, or expertise. When feeds start to fill with generic AI writing, the platform’s core value proposition comes under pressure.

More Than a Report Button

LinkedIn’s response is not limited to user reports. Srinivasan said the company is ramping up new classifiers to identify whether a post is “AI-slop or generally low-quality content.” The intended result is to reduce the amount of such material shown in suggested posts and in content from outside a user’s network.

That language is important. It points to distribution, not only moderation. LinkedIn does not necessarily need to remove every weak or AI-heavy post to change user experience. If it can reduce amplification in recommendation systems, it can make feeds feel more relevant without turning every borderline case into a takedown decision.

The company is also testing a way to show people that other users feel their posts come across as inauthentic or rely heavily on AI. If expanded, that feature would move the platform beyond invisible algorithmic ranking and toward visible social feedback. Such a step could change posting behavior, particularly on a network where perception and credibility are closely tied to career outcomes.

At the same time, LinkedIn is removing a feature that used AI to “enhance” posts. It plans to replace that tool with one that proofreads a user’s words without changing their voice. That is a meaningful product shift. Instead of helping users generate more polished platform-native content, LinkedIn says it wants to preserve original expression while still offering lightweight assistance.

The Larger Platform Problem

LinkedIn’s move reflects a broader tension facing digital platforms in the generative AI era. AI tools can dramatically lower the cost of publishing, but they can also lower the average informational value of what gets published. If recommendation systems are not adjusted quickly, feeds can become saturated with repetitive, plausible-sounding material that is technically readable but socially empty.

Professional platforms may feel that pressure earlier than others because users are often less interested in entertainment than in signal: hiring trends, market shifts, technical expertise, project updates, and real operational lessons. Content that looks polished but says very little is especially corrosive in that setting.

LinkedIn is effectively acknowledging that content authenticity is now a product and ranking issue, not just a cultural complaint. By creating a report category with unusually plain language, expanding automated classifiers, experimenting with feedback signals, and removing at least one AI-writing feature, the company is trying to reposition itself against the flood of synthetic professional posting.

Whether that effort succeeds will depend on execution. User reports can be noisy. Automated systems can misread style as spam. Some legitimate users will continue to rely on AI drafting tools, and LinkedIn will need to distinguish between editing assistance and content that feels mass-produced or deceptive. But the strategic direction is clear: the company believes feed quality is being damaged by AI-generated filler, and it is now willing to say so in product terms that users immediately understand.

That makes this rollout more than a quirky new menu option. It is an early example of a major platform building explicit anti-slop controls into its interface, ranking systems, and writing tools at the same time. For users, it offers a direct way to push back on inauthentic content. For the platform, it is a test of whether social networks can preserve trust as generative text becomes cheaper and more pervasive.

This article is based on reporting by The Verge. Read the original article.