How to make an editorial political explainer video with AI
Vertical 9:16 explainer for one political topic — Wikipedia portraits, accurate US/state maps, kinetic captions, full-screen veil overlays for emphasis. ~55s, ready to post.
why this format works
Political explainers work when one specific question gets a clean factual answer in under a minute. The dominant visual language for the genre on TikTok, Reels, and Shorts is editorial: bold sans-serif captions, a real map, real portraits of the named politicians, and a paper-grid background that says editorial not influencer. Use this for explaining a single race, a single bill, a single procedural quirk ("why does Georgia hold runoffs?"). Use a different format if the topic is partisan opinion or an open editorial — explainers earn trust by being scrupulously factual.
what ivy does
What's automated, scene by scene.
Reads the article or transcript
Paste a news article URL or a transcript and ivy pulls the named entities (politicians, states, dates), the central question, and the factual answer. If the topic isn't yet covered in a single article, you can describe it in chat and ivy will research and verify.
Sources real Wikipedia portraits
For every named politician, ivy fetches their Wikipedia portrait and frames it in a portrait bubble with their name and party. Wikimedia-licensed only — no AI faces, no stock photos, no scraped editorial images.
Builds accurate US and state maps
Real geography from us-atlas TopoJSON. State outlines, county-level fills, and highlight overlays for the specific district or county the explainer is about. Hand-drawn state shapes are a tell that the video is a knockoff; ivy uses the same datasets cartographers do.
Writes editorial kinetic captions, not voiceover-only
Captions advance mid-scene, in time with the voiceover. Single short factual phrases — chyron-length, not paragraph-length. The full transcript is generated by Whisper from the voiceover, then segmented into caption phrases that fit the screen.
Uses full-frame veil overlays for emphasis
When a single fact needs to land hard ("50% OF THE VOTE," "ZERO BLACK SENATORS UNTIL 1967"), the scene veils with a near-black overlay and the fact lands in the center as floating typography. The visual vocabulary that makes editorial explainer shorts feel authoritative.
Returns a brief you can edit before rendering
Swap a portrait, re-time a beat, change the highlighted county, soften the caption. After the brief, render is ~5 minutes for a 55s explainer.
make one in five minutes
Six steps. Five minutes.
Pick a single, factual question
"Why does Georgia hold runoffs?" "What's the Senate parliamentarian?" "How does redistricting work?" One question, one answer. Vague topics produce vague explainers.
Paste a source article or transcript
An article URL or pasted transcript anchors the facts. If you don't have one, describe the topic in chat and ivy will research and propose sourced facts — you'll see citations in the brief.
Confirm portraits and map highlights
Brief shows: "Will use Wikipedia portrait of [Politician X], highlight Georgia in the US map, county-level fills for the runoff race." Override if a portrait is wrong or a map needs different highlighting.
Press Render
Vertical 9:16 by default. Ivy returns an MP4 in ~5 minutes. About 55-65 seconds.
Refine via chat
"The third caption is wrong, change it to: ..." "Use a different photo for Senator X." "Add a final card with the date the runoff happens."
Post to TikTok / Reels / Shorts
Download the MP4 or post directly to X from inside ivy. For TikTok and Instagram, download and upload there.
what to watch out for
Hard-won lessons from shipping these.
Stay factual, not editorial
Explainers earn their reach by being scrupulously factual. The moment an explainer slips into opinion, comments turn against it and the algorithm punishes the engagement signal. If you want to argue a position, use a different format (hot-take or commentary). Explainers answer questions; they don't argue.
Maps must be accurate
Hand-drawn US/state shapes immediately read as low-quality. Ivy uses us-atlas TopoJSON for accurate geography. If you need a map of a specific congressional district or county, name it in the prompt — ivy will use real FIPS codes to highlight the right region.
Real portraits, real attribution
Every named politician shows up as a Wikipedia portrait. If the politician isn't on Wikipedia (rare for federal/state-level), the brief will flag it and ask you to upload a public-domain photo. AI-generated portraits of real politicians are out of bounds.
Caption pacing
Captions advance mid-scene with the voiceover, not as paragraph blocks. Each caption is one short factual phrase that fits in roughly two lines. Long captions read as bad subtitles and tank retention. Default ivy caption length is good — only override if you have a specific reason.
Veil overlays for emphasis, used sparingly
Full-frame dark veils with a giant fact in the middle are the format's signature, but they only work if you use 1-2 per video. Overuse them and they stop landing. Reserve for facts that need to feel like a punch.
What ivy won't do
Won't fabricate vote totals, dates, or named individuals. Won't show a candidate the source article doesn't name. Won't produce explainers where the central premise is a contested partisan claim — if your prompt asks for one, ivy will push back and ask for a sourced article.
try one of these prompts
Steal these.
faq
Common questions.
Procedural and factual topics work best (how runoffs work, what a parliamentarian does, redistricting mechanics). Contested partisan claims are out of scope — ivy will push back and ask for a sourced article.
Your first one's on us.
Paste a URL, get a video back in three minutes. Two free credits to start.