=== AI Blackout - Block AI Crawlers & Stop AI Training === Contributors: lazycodelab, adictonat0r Donate link: https://paypal.me/lazycodelab Tags: ai, ai crawlers, robots.txt, content protection, copyright Requires at least: 6.8 Tested up to: 7.1 Requires PHP: 7.4 Stable tag: 2.0.0 License: GPLv2 or later License URI: https://www.gnu.org/licenses/gpl-2.0.html Stop AI companies scraping your posts to train their chatbots. Block AI crawlers, state your terms, and switch off the AI inside WordPress. == Description == **AI companies are copying your content to train their chatbots. Nobody asked you, and nobody paid you.** A crawler takes your writing. Later someone asks a chatbot a question, gets an answer built out of your work, and never visits your site. You lose the traffic, the ad view, the sale, the signup. Most site owners have no idea it is happening. No email, no warning, nothing in your dashboard. **AI Blackout shuts the door.** Install it, answer one question, done. = What you get = * **26 AI crawlers blocked.** OpenAI, Anthropic, Google, Meta, ByteDance, Perplexity, Amazon, Apple, and the dataset scrapers that feed everyone else. * **Your refusal on the record.** A formal legal notice that your work is not free to train on, published where AI companies look for it. * **AI switched off inside WordPress.** Core's AI, the AI your plugins keep adding, and anything on your site trying to reach an AI company. * **Google traffic untouched.** Search engines are protected by design, not by you being careful. * **Proof it works.** The plugin asks your own site for a page while pretending to be a crawler, then shows you exactly what it got. = Setup is one question = Simply pick one: * **Block AI from taking my content.** Refuses training and scraping crawlers, leaves AI search alone so people can still find you. *Most people want this one.* * **Block everything AI.** Also refuses the crawlers behind ChatGPT, Claude and Perplexity search. Ordinary search engines are still allowed. * **Just state my terms.** Blocks nobody. Publishes your legal position and asks, in robots.txt, that your work is not used for training. Press Save. The plugin writes every layer your server can take. No apply button, no second screen. Change your answer later and it rewrites everything to match. = Your Google ranking is safe = Getting dropped from Google is the worst thing a plugin like this could do to you, so it is prevented by design rather than by good intentions. Googlebot and its seven siblings, Bingbot, Yahoo, DuckDuckGo, Baidu, Yandex, Applebot and the Facebook, X/Twitter and LinkedIn preview bots sit on a protected list of 18 identifiers. Every rule is tested against that list before a byte is written, and anything that would catch one of them is thrown away. A few AI crawlers do affect whether you show up in AI answers. Those are flagged on their own row, and blocking one takes a separate confirmation. = Three layers of blocking = The plugin tells you which ones are actually live on your server instead of assuming all three are. 1. **robots.txt.** A polite request. Honest crawlers obey it. 2. **WordPress.** Your site refuses the request itself. 3. **Your web server.** On Apache and LiteSpeed, crawlers are turned away before WordPress or your cache is involved. This is the strongest layer. = Your rights, in writing = Blocking stops the well-behaved crawlers. A declaration removes everyone else's excuse. * A reservation of rights under Article 4(3) of the EU copyright directive, at the address AI companies check. * Content Signals in robots.txt, separating "you may index this" from "you may train on this". * Meta tags for the crawlers that read those instead. * A matching clause for your terms page, written for you. It keeps a dated record of every wording you have published, so you can answer "since when?" about your own site. One click sends your robots.txt to the Internet Archive for a timestamped copy. = Getting AI out of your WordPress = WordPress 7.0 shipped with AI in core, and your plugins keep adding more of it. * Turn off core's AI and close the interface that lets AI agents act on your site. * Hide the screen where AI provider keys get entered. * Strip the AI out of Elementor and Yoast SEO using their own off-switches. Nothing is written into their settings. * Log every request your site makes to an AI company: which plugin, where to, how often. It watches by default and blocks only when you say so. = What it cannot do = This part matters more than the feature list, so please read it. * **It cannot stop a crawler pretending to be an ordinary browser.** Nothing running on your server can. Blocking works because honest crawlers say who they are. * **robots.txt is a request, not a barrier.** A crawler that ignores it is unaffected. This plugin never counts robots.txt as protection or shows it to you as though it were. * **On nginx it cannot write server rules.** nginx config lives outside your site folder and needs root and a reload, so no WordPress plugin can touch it. This one writes the rule out for you with the message to send your host. Any plugin claiming to block crawlers at server level on nginx is not telling you the truth. * **A declaration is a legal position, not a lock.** It does not physically stop anyone. It removes their excuse. = Nothing is sent anywhere = The crawler list ships inside the plugin and updates when the plugin does. **AI Blackout makes no external requests of any kind.** No account, no licence key, no analytics, no remote ruleset, no "anonymous usage data". The only request it ever makes is to your own site, when you press **Test it from the outside**. = Your files stay yours = The plugin writes `robots.txt` and, on Apache or LiteSpeed, `.htaccess`. It writes them when you save, so what is on disk always matches what the screen says. Everything goes between its own markers and nothing outside them is touched. Removing it puts the file back byte for byte, and tells you plainly if something else edited the file in the meantime. **Deactivating the plugin removes the rules it wrote.** A blocking rule with no plugin managing it is a booby trap, and it would make "deactivate this and see if it was the cause" give you the wrong answer. Your settings and your dated record stay exactly where they are. = Not affiliated with anyone named here = AI Blackout is an independent plugin by LazyCodeLab, not affiliated with or endorsed by any company it names. OpenAI, Anthropic, Google, Elementor, Yoast SEO and the rest appear only to identify whose crawler is being blocked or whose AI feature is being switched off. All marks remain the property of their owners. == Installation == 1. Install and activate the plugin. 2. Go to **AI Blackout > Your content** and answer the one question at the top of **Protection**. Press Save. 3. That is genuinely it. You are protected. To see what is actually working, open **Advanced > Layer by layer**. It breaks your protection down one layer at a time and tests it against your live site rather than assuming. == Frequently Asked Questions == = Will this hurt my Google ranking? = No. Search engines sit on a protected list of 18 identifiers, and every rule is tested against that list before it is written. A rule that would catch one of them is thrown away. If you choose to block an AI crawler that affects whether you appear in AI assistants' answers, the plugin says so plainly and asks you to confirm that one separately. = Do I need to know anything technical? = No. Install it, answer one question, press Save. The plugin works out what your server supports and writes the right things in the right places. You never have to touch robots.txt or `.htaccess` yourself, and you cannot break your site by picking the wrong answer. = Does it slow my site down? = No. Nothing is loaded on the front end: no scripts, no stylesheets, no per-page database queries. Blocked crawlers are turned away by a user-agent check and nothing else. = Can I still let AI assistants link to my site? = Yes, and that is the default. Crawlers that collect training data get refused. The ones that fetch a page because a real person asked a question are allowed through. Every crawler is listed separately with what it actually does, so you can go through them one at a time if you want to. = Will it break Elementor or Yoast? = No. Their AI features are switched off using the off-switches those plugins ship themselves, and nothing is written into their settings. Turn it off here and everything is exactly as it was. = I am on nginx. Is it still worth installing? = Yes. robots.txt and the WordPress-level refusal both work on nginx with no setup. The only thing you cannot have without your host's help is the server-level rule, and the plugin writes that rule out for you along with the message to send them. = Does it work with a caching plugin? = The robots.txt and server-level layers do, because they run before your cache. The WordPress-level refusal does not, because a page served from cache never reaches WordPress at all. The plugin detects your cache and tells you which layers that affects, rather than leaving you to work it out. = Does it delete AI API keys I have already saved? = No. Existing keys are left alone. They simply cannot be used while AI is switched off. Keys set in `wp-config.php` or your server environment cannot be touched by any plugin at all. This one names them so you at least know they are there. = What happens to my files if I deactivate or delete the plugin? = They go back to how they were. Deactivating removes the rules byte for byte, and reactivating writes them again from your settings. Deleting removes the lot by default: rules, settings, the activity log table, the scheduled cleanup job. **No leftovers, no orphaned database rows, no mystery lines in your `.htaccess` six months later.** There is a checkbox on the **Legal declaration** tab that keeps your settings and dated record through a delete. Worth ticking if you have been publishing a reservation of rights for a while, because that record is the only proof of when yours started and it cannot be recreated later. It keeps information, not protection: once the plugin is gone, nothing is blocking anyone either way. = Why is it called AI Blackout? = Because that is literally what it does. The site goes dark to AI. It cannot get in through your crawlers, and it cannot run inside your WordPress. Both halves are the same idea. = Enjoying it? = A review is the biggest thing that helps another site owner find this before their content gets scraped. If it is working for you, [leaving one](https://wordpress.org/support/plugin/ai-blackout/reviews/#new-post) takes thirty seconds. == Screenshots == 1. Protection: what is actually protecting you right now, and how long it has held without a break. 2. One question decides everything. Saving writes every layer your server can take, with nothing left to apply afterwards. 3. AI inside WordPress: switch off the AI built into core, close the AI agent interface, and strip the AI your other plugins added. 4. Advanced: every AI crawler, what it does, and your decision. Anything that would cost you visitors says so on its own row. 5. Advanced: layer by layer, tested by asking your own site for a page the way an AI crawler would, rather than assumed. 6. Legal declaration: answer use by use, in the wording the Content Signals policy defines, published in your robots.txt. 7. Your dated record of what you have declared and since when, plus a one-click Internet Archive snapshot of your robots.txt. 8. Where your protection stands without opening the plugin: a dashboard widget and an admin bar entry. == Changelog == = 2.0.0 = *2026-09-11* **Added** * What's new screen, shown once after each update. * Dashboard widget and admin bar entry showing current protection state. * Status line for the plugin on the Plugins screen. * Running record: days protected without a break, how long rights have been reserved, total requests refused. * One-click Internet Archive snapshot of your robots.txt. * Crawlers that also bring you visitors are flagged in the crawler list, with Allow marked as recommended. **Changed** * **Breaking:** enforcement now applies automatically on save. The `.htaccess` rules that stayed off in 1.0.0 until you previewed and confirmed them are now written to match your settings. * Status, Setup, Crawler Control and Declarations merged into one screen with three tabs: Protection, Legal declaration, Advanced. AI inside WordPress is unchanged. * Reactivating the plugin restores its rules automatically instead of asking first. * Anything needing your attention moved to the top of Protection. * The dated record leads with the date and is no longer described as evidence. **Removed** * Per-layer on/off switches and the apply and revert buttons. * The notice prompting you to finish setup. * The `ai-blackout-setup`, `ai-blackout-crawlers` and `ai-blackout-declarations` pages. There is no redirect, so update old bookmarks to `ai-blackout`. **Fixed** * The declaration record added three rows on every save, because it compared each entry against the previous entry of any type rather than its own. It now records only real changes and clears duplicates left behind by earlier versions. = 1.0.0 = *2026-08-28* * Initial release. == Upgrade Notice == = 2.0.0 = Saving now writes your server rules for you. The .htaccess rules that stayed off until you confirmed them in 1.0.0 are applied automatically from 2.0.0, and the per-layer switches are gone. Check Advanced after updating. Your crawler choices, declarations and dated record are untouched. = 1.0.0 = First release, so nothing to upgrade from yet. If you are installing it fresh, please read the "what it cannot do" section before you trust it with anything. That part is the honest bit.