LinkedIn is rolling out a button that lets users flag posts as "seems like AI slop," marking the latest push by a major social platform to fight back against the rising tide of low-quality, artificially generated content flooding the web. The Microsoft-owned professional network announced the feature Thursday as part of a broader crackdown on automated posts, comments, and engagement, as TechCrunch reported.
The decision reflects a wider reckoning across the internet. Platforms from Substack to Meta are scrambling to build tools that detect and demote AI-written material, as users grow tired of scrolling past computer-generated fluff that drowns out authentic voices. LinkedIn's move — confirmed in a LinkedIn post by chief product officer Hari Srinivasan — is one of the most direct yet.
How LinkedIn's new system works
Srinivasan said the "seems like AI slop" button is just one piece of a larger effort. The company is now blocking "hundreds of thousands" of automated comment attempts each day, and millions of automation attempts in total over the past couple of months. New classifiers have been deployed to detect AI-generated or low-quality posts before they reach users' feeds, particularly in suggested content from outside a user's network.
The reporting button feeds directly into those classifiers. Each time someone flags a post as slop, LinkedIn uses that signal to train its moderation models to get better at spotting the stuff on their own. The company will also start privately notifying users when their own posts are getting flagged as inauthentic due to heavy AI use — in the hope that people will dial back the automation rather than risk looking fake.
Perhaps the most telling change: LinkedIn is removing its own "enhance your post" AI feature, which helped users rewrite their updates. It's being replaced with a tool that proofreads without altering the user's voice — a quiet admission that even the platform's own AI writing assistant may have been part of the problem. Users who relied on that feature to polish their professional writing will need to adjust, but LinkedIn argues the trade-off is worth it if it means fewer robotic-sounding posts in the feed.
A platform-wide problem
LinkedIn isn't alone in this fight. Last week, newsletter platform Substack added a tool to flag AI-written content through a partnership with AI-detection startup Pangram, as TechCrunch reported. That same week, Pangram announced $9 million in fresh funding to expand its detection technology across more platforms, covered by TechCrunch. The funding round signals that investors see AI-content moderation as a growth market, not a passing concern.
The scale of the issue is sobering. Internet infrastructure firm Cloudflare reported this month that bot traffic has now overtaken human-generated requests on the web — a milestone the company reached faster than it had predicted, according to NBC News. Cloudflare CEO Matthew Prince noted on X that the crossover happened sooner than the company had anticipated, underscoring how quickly automated traffic has grown. Digg's attempt to relaunch as a Reddit competitor collapsed in March partly because its team couldn't get a handle on bots flooding the site.
Substack's partnership with Pangram gives newsletter readers a transparency indicator showing whether content was written with AI assistance. The system analyzes writing patterns rather than just checking for obvious bot markers, making it harder for sophisticated AI content to slip through. Pangram's approach relies on statistical analysis of sentence structure, vocabulary diversity, and repetition patterns — the same tells that human readers pick up on when a post feels off.
The bigger picture: AI racing ahead of oversight
LinkedIn's policy shift arrives as unease grows inside the AI industry itself. More than 1,100 employees and senior researchers from OpenAI, Google, Meta, Microsoft, Amazon, and Anthropic signed an open letter this week urging the US government to prepare safeguards that could slow the pace of automated AI development if needed, as the Times of India reported.
"AI could help create a dramatically better future, but that outcome is not guaranteed," the letter states. "There is a real risk that capability development rapidly accelerates beyond our ability to understand or control the resulting systems."
The signatories — including Anthropic CEO Dario Amodei, OpenAI chief scientist Jakub Pachock, and Anthropic chief science officer Jared Kaplan — called for international coordination on technical and governance tools for what they describe as "automated AI development." They asked the US government to support an international effort to pace the frontier of AI advancement deliberately. The letter specifically flags the danger of AI systems that can improve themselves without human oversight, a scenario that could rapidly outstrip existing safety measures.
What this means for the content ecosystem
The simultaneous push from platforms like LinkedIn and warnings from AI insiders points to a common anxiety: the tools that generate content are getting cheaper and better faster than the tools that filter it. Months ago, AI-written posts on LinkedIn were easy to spot — generic platitudes, odd phrasing, suspiciously perfect formatting. Today the line is blurrier. Language models can now produce first drafts that read naturally enough to pass a casual skim.
LinkedIn's approach is notable for its layered strategy. Rather than just hiding AI slop after the fact, the company is trying to discourage its creation in the first place. The private dashboard flags tell users when their content reads as inauthentic, giving them a chance to course-correct before their reach gets cut. Removing the "enhance your post" feature removes a tool that was effectively training users to sound more robotic. It is a rare case of a platform voluntarily giving up an AI feature to preserve the quality of its ecosystem.
For businesses and professionals who rely on LinkedIn for networking and lead generation, the changes cut both ways. Authentic voices should get more visibility as AI sludge gets filtered out. But anyone who uses AI as a writing aid — to fix grammar, rephrase an awkward sentence, or translate between languages — may find their posts flagged unfairly. The nuance between AI-assisted and AI-generated is still hard for algorithms to nail, and false positives could frustrate legitimate users.
What's next
LinkedIn said it will expand access to profile and page verification tools and add options to block comments from company pages users no longer want to see. The company declined to say how many AI slop reports it expects to process daily, but the volume of automated comment blocking — hundreds of thousands per day — suggests the problem is already massive. The new systems will roll out gradually, Srinivasan said, as the company tunes its classifiers to reduce false positives.
Srinivasan put it plainly: "AI slop is a top priority for all of us. People come to LinkedIn to connect with real people and share their real perspectives."
Whether the reporting button will make a real dent — or simply give users a satisfying way to vent — remains to be seen. But the direction is clear across the industry: platforms that spent the last two years racing to add AI features are now racing to contain the mess those features helped create.
The move also draws attention to the growing tension between AI-powered productivity tools and the human authenticity that social platforms were built on. LinkedIn's decision to remove its own "enhance your post" AI writer is perhaps the strongest signal yet that even the companies selling AI tools are starting to worry about what happens when those tools get used at scale. For more AI industry coverage, check out AI.