LinkedIn AI detectors, feed blockers and the slop button, compared.

If you search for a way to stop reading AI-written LinkedIn posts you will find four different kinds of answer, and they do four different things. This page sorts them: what each approach does to a post, where your text goes, what it costs, and which one fits what you actually want. We make one of the tools below and say so where it appears. Product details are taken from each tool's own listing on 21 September 2026 and may change.

The four approaches

1. Report it: LinkedIn's "Seems like AI slop" button

Added on 30 July 2026 to the three-dot menu on every post. Clicking it removes that one post from your feed and sends LinkedIn a signal it uses to tune its classifier and decide reach. It is the only option that can affect the author, and the only one that does nothing until after you have read the post. LinkedIn says the option was used more than a million times in its first two weeks and that views of what it classifies as slop fell 40%. Full detail: what the button does and what it doesn't.

Pick it as a civic duty on the obvious ones. It will not clean your feed.

2. Label it: AI detectors that annotate the feed

• Pangram Feed Scanner labels AI-written text on X, LinkedIn, Substack, Medium and Reddit, and anywhere in Chrome from a right-click. Labels only; nothing is hidden. Pangram is a commercial detection company and publishes accuracy studies.
• LinkedIn AI Filter colours posts red, amber or green, lets readers vote on whether a post feels generated, and adds a "Filter AI" toggle that hides the suspicious ones.
• LinkedIn Post AI Detector runs each post through Quillbot's AI checker and marks it.
• LinkedIn AI Slop Detector is the Firefox option: linguistic patterns, a "slop rating" label per post.

Pick one if you want to keep reading everything and just know. The cost is that the read still happens; the label arrives with the post, not instead of it.

3. Block it: feed blockers and cleaners

• Feed Blocker for LinkedIn and linkedin-feed-blocker remove the home feed entirely and leave jobs, messages and profiles alone.
• LinkedIn Feed Hider lets you choose which feeds vanish: home, company pages, profile activity.
• LinkedIn Feed Cleaner strips suggested and promoted posts and keeps the rest.

Pick one if the feed is a distraction you want gone. Nothing here reads the writing; the human posts go with the generated ones.

4. Filter it: score every post, hide the ones over the line

• LinkedIn AI Content Detector uses structural rules in the browser, three sensitivity levels, and blurs matches until you click.
• LinkedIn Feed Filter is open source, rule based, and also targets life-lesson posts, context-free selfies and recycled news. No text leaves the browser.
• No AI LinkedIn Feed sends each post to OpenAI, Anthropic, Gemini or a local model with your own API key and hides above a threshold from 1 to 100.
• NoSlop for LinkedIn, ours. Eleven rules on your device by default, no network calls. Opt in to Slop Score and every post gets a calibrated "% AI" badge from Jev, TypeSafe AI's decision model, through a proxy we run, on four dimensions: AI-written, engagement bait, fake story, recruiter or sales spam. You pick which of the four should hide a post and how strict to be. Every collapsed card lists its reasons and reopens with one click. Free, no account, no key. Chrome and Edge, LinkedIn only.

Pick one if you want the feed itself to change: fewer generated posts in front of you, the human ones untouched, and a way back when the score is wrong.

Two extensions often listed alongside these, AIBlock and ZenFeed AI, hide posts about AI, the topic, not posts written by AI. Different job.

What to check before you install any of them

• Where the text goes. In the browser, to the vendor, or to an API you pay for. Every tool above says which in its listing or privacy policy; if it doesn't, skip it.
• Hide or label. Decide whether you want to stop reading slop or to know that you read it.
• Can you see why. A score with reasons can be argued with. A verdict cannot.
• Your threshold, not theirs. Detectors are probabilities. You should set the line and be able to move it after a week of use.
• Undo. Anything that hides should reopen in one click, on that post, without turning the tool off.
• Effect on the author. Extensions change only your browser; nobody is notified and reach is untouched. LinkedIn's button is the exception, by design.

Where NoSlop sits, honestly

The default mode is private and blunt: rules caught 58% of generated posts in our 24-post calibration set with no false positives. Slop Score is accurate on that set, 92% accuracy and 83% recall, but the post text does leave your browser to our proxy and on to Jev, which is why it is off until you switch it on. It is Chrome and Edge only, desktop only, LinkedIn only. The numbers, the cost and the failures are in the technical write-up.

Frequently asked

What is the best AI detector for LinkedIn?

It depends on what you want done with the result. If you want labels you can read across many sites, a label-only detector such as Pangram Feed Scanner fits. If you want the feed cleaned rather than annotated, pick a filter that scores each post and hides the ones over a threshold you set, with a one-click undo. NoSlop for LinkedIn is the one we make: on-device rules by default, an optional calibrated percent-AI score from Jev, reasons on every hidden card.

Do LinkedIn AI detector extensions send my feed to a server?

Some do and some don't, so check before installing. Rule-based extensions (LinkedIn AI Content Detector, LinkedIn Feed Filter, NoSlop's default mode) analyse the text in your browser. Model-based ones send post text somewhere: to the vendor (Pangram, Quillbot via LinkedIn Post AI Detector, NoSlop's opt-in Slop Score through its own proxy) or to an API you pay for yourself (No AI LinkedIn Feed). None of them should need your LinkedIn login.

Is LinkedIn's "Seems like AI slop" button an AI detector?

No. It is a report. You are the detector: you read the post, decide it seems generated, and tell LinkedIn. The post leaves your feed and LinkedIn uses the signal to tune its own classifier and the post's reach. It gives you no score and does nothing to the next post.

Can I block AI-generated posts on LinkedIn without an extension?

Not by the writing itself. LinkedIn offers per-person controls (Unfollow, Mute, "I don't want to see this"), the Top/Recent sort, and the per-post "Seems like AI slop" report. None of them looks at the text before you do. A browser extension is currently the only way to filter the feed by how a post is written.

A tool we missed or described wrongly: hello@noslop.lol and we will fix it. What NoSlop does with data is on the privacy page.