All activities Professional learning
Teams race to verify viral claims and an AI-written summary by leaving the page, not staring harder at it.
Most of us were taught to judge a source by reading it closely: the About page, the spelling, the look of the site. Professional fact-checkers do the opposite. They leave the page and read laterally, checking what others say about the source before trusting it. In this relay, educators experience both approaches on a set of claims, including one confident AI-generated summary, and then redesign a lesson so students practice the move that works.
Hook
Flip Card 1 on the screen. Ask: "Trust it, doubt it, or need more?" Everyone holds up one, two, or three fingers. Don't reveal anything. Say: "Hold on to that gut feeling. By the end, you'll know what a fact-checker does in the next 60 seconds that most of us never do."
Facilitator noteMost rooms split. That split is the point: gut feelings about credible-looking content are unreliable.
Explore
Teams draw Cards 2–4 and may use only what is on the card: tone, formatting, the source's own description of itself. They record a verdict and a confidence level (1–5) on Handout B. No devices yet.
Facilitator noteListen for reasons like "it looks official" or "it has a .org address." Write a few on the chart paper, untouched. You'll return to them.
Model
Model SIFT on Card 1 out loud. Stop: notice your reaction. Investigate the source: open a new tab and search the organization's name plus words like "funding" or "who runs". Find better coverage: search the claim itself and see whether reliable outlets report it. Trace: follow the quote or number back to its original context. Narrate every click, including dead ends.
Facilitator noteUnplugged: read the "Other sources" key text aloud as though it were your search results. The thinking move is identical.
Practice
Teams re-verify Cards 2–4 and take on Cards 5–8. Roles rotate every card: Searcher (drives the device or asks for the key), Skeptic (asks "who's behind this?"), Scribe (logs moves on Handout B). Card 7 is an AI chatbot's summary. Teams must trace each of its three factual claims before calling it trustworthy.
Facilitator noteTeams usually find that Card 7 is fluent, confident, and two-thirds right. The invented statistic is the teaching moment: a polished tone is not evidence.
Debrief
Compare Round 1 and Round 2 confidence scores. Ask: "Which cards flipped? What move flipped them?" Add the moves to the "Moves that worked" chart beside the Round 1 reasons. Then ask about Card 7: "How is checking an AI summary like checking a website, and how is it different?"
Facilitator noteKey difference: an AI summary has no author page to investigate, so tracing claims is the only move available. Students need that habit.
Transfer
Each person uses Handout C to pick an upcoming lesson where students use online or AI-generated information, and rewrites the task so students must turn in verification evidence (a move, a source, a change they made), not just the final answer. Pairs swap and ask: "What evidence will show that students verified?"
Facilitator noteThis is the 2026 ELE question in action. The technology matters only if it makes students' verification visible.
Run both rounds on paper. Each card's facilitator key includes an "Other sources" note: what a lateral search would turn up. When a team asks "what do others say about this source?", hand them that slip. The cognitive move of leaving the page and checking elsewhere is preserved, and asking for the slip becomes a visible habit.
Give each team one device for Round 2 only, so Round 1 stays a pure vertical read. In an LMS or virtual session, post the cards as a slide deck, run the relay in breakout rooms, and have scribes log moves in a shared document using Handout B's columns. The facilitator projects any AI chatbot demonstration; participants don't need accounts.
Does it need a screen? The live search is where people discover how fast the lateral move is: often 30 seconds to learn who funds a site. That speed is what convinces them it is worth teaching. The paper version teaches the same move but can't show how cheap it is.
What you should be able to see or collect if it worked.
Teachers use SIFT to verify claims, including an AI summary, preparing students to judge source reliability and show verification evidence.
Run Round 2 with students using the grade-appropriate cards, then collect Handout B as an exit ticket. Bring three samples to your coach or PLC and compare how many students named a verification move.
