BaitBlock

Protects you against phishing across the web — pauses credential-stealing sites and flags fake emails with on-device AI.

Pauses credential-stealing sites before you type your password — and flags phishing emails in Gmail. 100% local, on-device AI. Description: Phishing, caught before you bite. 🎣 The web is full of fake emails and sites pretending to be something else, and when you lose attention, you too can fall into a trap and enter your password or credit card information. BaitBlock is a Chrome extension that checks every page you visit to verify it's not a phishing scam. It uses an on-device AI model which costs you nothing — seriously, it's free and will always protect you, your parents, your kids, and anyone you share this with. 🌐 ON EVERY SITE — when a page asks for a password, card number, one-time code, or ID document, BaitBlock checks whether the site is what it claims to be: lookalike domains, borrowed logos linking to a different organization, forms submitting your input elsewhere. A fake login page is paused behind a full-screen warning before you can type a single character. If you reach a form while a check is still running, BaitBlock holds your input until the verdict is in — or lets you knowingly continue. 📧 IN GMAIL — every email you open is scanned instantly: • Senders impersonating brands from lookalike domains ("alert.postguidelines-x.com" pretending to be X) • Links cloaked behind Google AMP cache, redirect services, or URL shorteners — the REAL destination is revealed next to each link • Link text showing one domain while pointing to another • Punycode/homoglyph domains, raw-IP links, abused TLDs, account-threat pressure language Suspicious emails get a warning banner explaining exactly why, and dangerous links are blocked until you explicitly unlock them. 🤖 AI THAT REASONS, FULLY ON-DEVICE: BaitBlock uses Chrome's built-in AI (Gemini Nano) to judge borderline pages the way a security analyst would — comparing the claimed identity against the actual domain, weighing the evidence, and explaining its verdict in plain language. On first install, Chrome downloads Google's on-device model (a few GB, one time, capable desktops only); after that, AI analysis works offline. On devices without it, BaitBlock runs on its instant rule engine alone. 🚦 ALWAYS VISIBLE, NEVER MYSTERIOUS: • The toolbar shield is a live status light — green when clear, blinking yellow while checking, red with a count when something was flagged • Open the popup for an instant verdict on the current site: "Passed — not phishing", a warning, or blocked • Full history of everything flagged and exactly why, with the analysis time of every AI verdict ⚖️ PROPORTIONAL, NOT PARANOID: pages collecting real secrets can be paused outright; pages asking only for simple details get at most a dismissible note. Domains that pass a check are remembered (and listed in the popup, removable anytime) and never rescanned — zero overhead on the sites you use every day. 🔒 PRIVACY, ABSOLUTE: all analysis — including the AI — happens locally in your browser. No email content, no browsing history, no data of any kind ever leaves your machine. No accounts, no tracking, no external servers. The extension makes zero network requests.