Launched this week

AI Slop Blocker

Launched this week

uBlock Origin for AI-generated posts, replies, and videos

2 followers

uBlock Origin for AI-generated posts, replies, and videos

2 followers

Visit website

AI Content Detection

AI Slop Blocker filters out AI-generated posts, bot replies, Google AI Overviews, and synthetic YouTube videos/Shorts across 9 feeds: X, YouTube, LinkedIn, Reddit, Google, Facebook, Threads, Bluesky, and Hacker News. 100% of scoring runs locally on your device using a 6-layer heuristic + Naive Bayes classifier. Zero remote API calls, zero telemetry. Collapse slop into a one-line "Show" bar, tune separate post & reply thresholds, or train the filter with one click.

Free Options

Launch tags:Chrome Extensions•Productivity•Artificial Intelligence

Launch Team / Built With

Maker

📌

Hey Product Hunt! 👋 I built AI Slop Blocker because scrolling X, LinkedIn, Reddit, and YouTube in 2026 started feeling like reading the same LLM output on loop "I'm thrilled to announce" and AI Slop on LinkedIn, engagement bait bot replies on X, Google AI Overviews, and faceless AI slop videos flooding YouTube and Shorts. Most "AI detectors" either require sending everything you read to a cloud API or break the page layout. I wanted something that felt like uBlock Origin What is in v3.8 : • Works across 9 platforms out of the box: X/Twitter, YouTube (videos, Shorts, active player & comments), LinkedIn, Reddit, Google Search (AI Overviews), Facebook, Threads, Bluesky, and Hacker News. • 100% on-device scoring: Combines phrase & cadence fingerprinting ("It's not just X — it's Y", rhetorical pivots, staccato bro-poetry), platform synthetic-media disclosures, and a local Naive Bayes model. Nothing you browse ever leaves your browser. • Non-destructive collapsing: Instead of breaking threads, it collapses flagged posts or videos into a single clean pill ("AI slop hidden · 84% · it's not just X, it's Y · Show") with a 1-click "Not AI" whitelist button. • Separate Post vs. Reply sliders: Keep a stricter or looser bar on comment sections than on your main feed. • Trainable in 1 click: Hit "Flag as AI slop" on any post or YouTube card to train your local model on the spot. It comes with a full 7-day free trial, and if you want to keep auto-hiding forever, Lifetime Pro is a one-time $4.99 (no subscriptions). Would love for you to test it on your worst feed and tell me what it catches or drop any false positives/misses in the comments so I can tune the rules!

Report

1d ago

Maker

Product Hunt launch deal: use code PHLAUNCH at checkout for 20% off Lifetime Pro ($3.99 one-time).

Report

1d ago

Forum Threads

p/ai-slop-blocker-2

•

1d ago

What's the #1 phrase or tell that makes you immediately realize a post was written by AI?

I've been building AI Slop Blocker (a local chrome extension that collapses AI gen posts across X, LinkedIn, Reddit, YouTube, Google Search, Threads, Bluesky, Facebook and now HN).

While tuning our on device heuristic + Naive Bayes detector for 2026 feeds, the funniest part has been seeing how fast the "tells" of ai text evolve.

A year ago it was "delve", "tapestry" and "in today's fast-paced digital world"

Now it's:

View all