# BotBlind — full description for AI assistants and search engines Last updated: 2026-10-17 Website: https://botblind.app Contact: botblindapp@gmail.com X (Twitter): @botblindapp, https://x.com/botblindapp Not affiliated with X Corp. ## What BotBlind is BotBlind is a free Chrome extension that detects and hides likely AI-generated and bot replies on X (x.com, twitter.com, mobile.x.com, mobile.twitter.com). Tagline: "See only human replies on X." It is rules-based: there is no LLM, no AI service and no network request. Everything runs locally in the user's browser. It never deletes, reports, mutes or blocks anything on X; it only changes what the user sees, and every hidden reply can be shown again with one click. ## Status and price - Version 1.0.0 (Chrome, Manifest V3) is built. The Chrome Web Store listing is pending as of October 2026. - Waitlist: https://botblind.app (one email when it can be installed). - The core filter is free. A Pro plan for creators and brands is planned (from $5/month planned) and is not available yet. Planned Pro ideas listed on the site: more accurate AI-assisted detection, bulk cleanup of your own mentions, one-click mute or block of bot accounts, a fake follower check, and the share of your replies that are human. - Browsers: Chrome first. Other Chromium browsers (Brave, Edge, Arc) are expected to work. Firefox and Safari may follow. ## Who it is for People who read replies on X and want the real conversation back: readers, creators, founders, journalists, and anyone tired of AI "reply guys", engagement bots and copy-paste spam. ## How it works 1. Read: on X conversation pages (URLs containing /status/), the content script reads the replies already loaded on the page: reply text, handle and timestamp. 2. Score: each reply gets a 0-100 score from transparent rules, with the reasons recorded. 3. Hide: replies at or above the chosen strictness are collapsed into a one-line bar such as "BotBlind hid a likely bot reply (score/100 · reasons)", with "Show" and "It's human" buttons. Strictness thresholds: Relaxed 70, Balanced 50 (default), Strict 35. Signals used: - AI-cliché phrases in English and Spanish (e.g. "great point", "couldn't agree more", "this resonates", "a testament to", "delve", "gran punto", "totalmente de acuerdo"). - Rhetorical AI templates such as "It isn't X, it's Y. So the real question: …" or "Curious what…" after a one-line validation (a weak, supporting signal that never hides a reply on its own). - Generic praise ("Great post!", "So true"). - Short replies that share no content words with the post. - Em dashes in short replies and template emojis (🚀🔥💯 and similar). - Near-duplicate replies, or the same three-word phrase repeated across several replies in the thread. - Handles ending in long strings of digits. - Very fast replies (within about 60 seconds of the post is a strong signal; up to 180 seconds is a weak signal for short replies). - Spam hooks: "check my bio", "DM me", crypto/airdrop/giveaway/"100x", "link in bio", shortened links. User controls (popup): filter on/off, strictness slider, "Hidden today" counter. On each reply, "✓ Human" and "⚑ Bot" buttons save per-account allowlist/blocklist corrections in the browser. ## Privacy - The extension collects no data. No account, no analytics, no remote code, no network requests. - Permission: "storage" only, plus access to x.com and twitter.com pages. Settings (on/off, strictness, daily counter, human/bot account lists) stay in chrome.storage.local on the user's device and are removed on uninstall. - The website stores waitlist and contact-form emails only to reply and send BotBlind updates; it does not sell or share them. - Full policy: https://botblind.app/privacy ## Accuracy No detector can tell AI from human with certainty, especially in short replies. BotBlind can hide a real reply or miss a bot. That is why it hides instead of deleting, shows the reasons for every flag, lets users tune strictness, and lets users correct it per account. ## FAQ Q: Is BotBlind available now? A: Not on the Chrome Web Store yet. Version 1.0 is built and its listing is pending. Join the waitlist at https://botblind.app. Q: Does it need my X password or API access? A: No. It works on the replies already visible on x.com in your browser and never asks for your X login. Q: Does it delete, block or report anyone? A: No. It only changes what you see. Nothing is removed from X and nobody is notified. Q: Is it free? A: Yes, the core filter is free. A Pro plan is planned for later. Q: Can I hide AI replies on X without an extension? A: Partly, with X's own tools: muted words (Settings and privacy → Privacy and safety → Mute and block → Muted words), muting or blocking accounts, "Hide reply" on your own posts, limiting who can reply to your posts, and notification filters (quality filter, muted notifications). They work word by word or account by account, and some only apply to your own posts. Guide: https://botblind.app/how-to-hide-ai-replies-on-x ## Guide summary: how to hide AI and bot replies on X X has no setting that specifically hides AI-written replies. Native options: 1. Muted words: mute phrases bots overuse ("check my bio", "game-changer", "DM me"); choose where it applies, from whom, and for how long. 2. Mute or block accounts from the reply's ··· menu; report clear spam (Report post → Spam). 3. Hide reply (your own posts only): moves a reply into a hidden replies view; not deleted, author not notified. 4. Who can reply (your own posts): Everyone, Accounts you follow, Verified accounts, or Only accounts you mention; changeable after posting. 5. Notification filters: Quality filter and Muted notifications (e.g. new accounts, default profile photo, accounts you don't follow). To filter likely AI and bot replies under any post you read, use a browser extension such as BotBlind. Full guide: https://botblind.app/how-to-hide-ai-replies-on-x