inLie — Political Manipulation Detector

See manipulation, propaganda & spin in political videos — live AI detection of rhetoric tactics, dodged questions and bias.

See the manipulation, not just the words. inLie is a live manipulation detector for political YouTube videos. While a debate, interview, or speech plays, AI analysis (Claude, on your own API key) names the rhetorical and psychological tactics as they're spoken — fear appeals, gaslighting, whataboutism, propaganda techniques, dodged questions, us-vs-them framing, and self-contradictions — each with the verbatim quote and a one-line insight into why the move works on you. FACT-CHECKING TELLS YOU WHAT WAS FALSE. INLIE SHOWS HOW YOU WERE STEERED. Fact-checks arrive days later and argue about claims. inLie works on a different layer, live: the engineering of the message itself — the spin, the framing, the pressure. It never rules on what's true; it shows you how the argument was built. Media literacy, applied in real time. WHAT YOU GET • A live ledger: every manipulation move, time-stamped and clickable — jump the video to the exact moment the words were said • A manipulation-density timeline: see where the video gets engineered • A per-speaker report: who dodged questions, each speaker's favourite tactics, and what their rhetoric was built to do • Clean moves credited in green — constructive rhetoric is named too • A shareable verdict: the money sentence, the tell, the contradictions • Whole-video scan: analyse everything up front in minutes — you see the estimated cost before a single token is billed WORKS ON Debates, interviews, speeches, panel shows, campaign ads, press conferences — any English-language political video on YouTube with captions. PRIVATE BY ARCHITECTURE No account. No tracking. No inLie server. Your Claude API key is stored on your device and sent only to Anthropic — analysis bills your own key (roughly $0.50–1 per hour of video), and inLie takes nothing. Your analysis history stays on your device. Full policy at https://ivan629.github.io/inlie-privacy/ HONEST BY DESIGN Every quote is verified verbatim against the transcript before it's shown. inLie labels how someone argues — never whether they're right. That discipline is the product.