Professional Learning  ›  EdTech & AI  ›  Workshops  ›  Deepfakes & Synthetic Media
a ready-to-run workshopDeepfakes & Synthetic Media
EdTech & AI · Workshop

Deepfakes & Synthetic Media (Workshop)

Everything you need to lead this with your staff or PLC: a minute-by-minute run-of-show, what to say, what to hand out, and how to make it stick. Pick the length that fits your slot.

Key takeaway

AI can now generate convincing fake images, audio, and video, which creates two dangers at once, students being deceived by fakes and students dismissing genuine evidence as possibly fake, and the response is not to teach a checklist of detection tricks that will soon be obsolete but to build the habits of provenance, verification, and reasoned judgment that hold up as the technology improves.

Format
PLC / team meeting
Length
45 minutes
Group size
Any (works 4-40)
Who can run it
Any teacher-leader
You will need
Slides (5)Printed handoutExample synthetic mediaA timer
Aligned to
Learning Forward: Rigorous ContentInTASC 8: Instructional StrategiesResearch on media literacy, provenance, and synthetic media
Share / assign

Why it works

AI can now produce synthetic media, images, audio, and video of people saying and doing things that never happened, that is convincing enough to fool most people most of the time, and the quality is improving fast. For students growing up in this environment, learning to navigate synthetic media is a core media-literacy skill, but it has to be taught in the right way, because the obvious approach, teaching students to spot the tell-tale signs of a fake, is a trap. The visual artifacts that give away a deepfake today, the odd hands, the strange blinking, the mismatched lighting, are exactly the flaws the next generation of tools fixes, so a lesson built on detection tricks is obsolete within months and, worse, teaches students a false confidence that they can tell real from fake by looking, which they increasingly cannot. The deeper problem has two sides that must be taught together. The first is the obvious danger: students being deceived by fakes, believing a fabricated video of a public figure, falling for a scam using a cloned voice, or being misled by a synthetic image presented as evidence. The second is subtler and in some ways more corrosive: the liar's dividend, where the mere existence of deepfakes lets people dismiss genuine, inconvenient evidence as possibly fake, so that real footage of real events gets waved away, and students slide into a cynical stance where nothing can be believed. A media literacy that only warns about fakes feeds this second problem, and the goal is to steer between both, neither gullible nor nihilistic. The approach that holds up is to shift the question from how does this image look to where did this come from and how do I verify it. That means teaching provenance and verification habits that do not depend on detecting artifacts: trace media to its original source, check whether reliable outlets are reporting the same thing, use reverse-image search and lateral reading to see what others say about it, consider who made it and why and whether it benefits someone, and treat a single unverified image or video, however convincing, as a claim to be checked rather than evidence to be believed. It also means building an understanding of how synthetic media is made and why it is persuasive, so students grasp the capability without over-relying on spotting its current flaws, and teaching the emerging tools of provenance such as content credentials and authentication standards. And it means naming the liar's dividend explicitly, so students learn that the right response to a world with deepfakes is better verification, not blanket disbelief. The workshop helps teachers understand synthetic media and its two dangers, and teach the provenance and verification habits, rather than the soon-obsolete detection tricks, that let students navigate it.

The research: Research on media literacy, provenance and authentication, and the liar's dividend.

The run-of-show

Choose your slot. The agenda, timings, and length update to match.

    The core activity: teach provenance, not detection tricks

    The heart of the session is teaching the verification habits that let students navigate synthetic media, rather than the soon-obsolete tricks of spotting a fake by looking.

    Ask where it came from, not how it looksYou cannot teach students to spot fakes by looking, because the tells disappear with each update. Provenance and verification habits, trace it to source, check who else reports it, ask who benefits, hold up as the technology improves.

    Facilitator notes

    Beware the liar's dividendThe existence of deepfakes lets people dismiss real, inconvenient evidence as possibly fake. Teaching students to verify rather than to disbelieve everything is what keeps them from sliding into corrosive cynicism.

    Adapt it

    Participant handout

    One page for every teacher. Print it, or save it as a PDF.

    K12 Academics · Professional Learning

    Deepfakes & Synthetic Media: provenance over detection

    Teach students to navigate synthetic media: shift from spotting fakes by looking to tracing provenance and verifying, and steer between deception and blanket disbelief.

    1. Why I will not teach detection-by-looking tricks:
    2. The provenance habit I will teach (trace to source):
    3. The verification habit I will teach (lateral reading, reverse search):
    4. How I teach the liar’s dividend without breeding cynicism:
    5. How students land neither gullible nor nihilistic:

    Make it stick

    Provenance habits outlast every detection trick. Close by setting the follow-up:

    Common questions

    Shouldn't I teach students the signs that give away a deepfake?
    It is tempting, but it is the one approach almost guaranteed to fail, so it is worth resisting. The visual artifacts that currently betray a fake, odd hands, unnatural blinking, strange lighting or edges, are precisely the flaws each new generation of tools eliminates, so a lesson built on them is obsolete within months. Worse, teaching detection-by-looking gives students a false confidence that they can tell real from fake with their eyes, which is becoming less true by the month, and a confidently wrong student is more dangerous than an uncertain one. The durable skill is not detection but verification: instead of asking whether an image looks fake, students learn to ask where it came from, whether reliable sources report the same thing, who made it and why, and to treat an unverified image or video as a claim to check rather than evidence to believe. Those habits do not depend on the technology's current flaws, so they keep working as it improves.
    What is the liar's dividend, and why does it matter for students?
    The liar's dividend is the way the mere existence of convincing deepfakes lets people dismiss genuine evidence as possibly fake, and it matters because it is the subtler and in some ways more corrosive of the two dangers. The obvious danger is that a fake fools someone; the liar's dividend is the opposite, that real footage of real events gets waved away with but that could be a deepfake, so that authentic, inconvenient evidence loses its power. For students, this is dangerous because a media-literacy message that only warns about fakes pushes them straight toward it: having learned that anything might be fake, they can slide into a cynical stance where nothing can be believed, which is as disabling as gullibility. That is why the goal is to steer between both dangers, teaching that the right response to a world with synthetic media is better verification, checking provenance and corroboration to establish what is real, not blanket disbelief that treats everything as equally doubtful. Students need to come away able to establish what is genuine, not convinced that nothing can be.
    How can students verify media if they can't tell what's real by looking?
    They verify the same way a good journalist or fact-checker does, by investigating the source and context rather than scrutinizing the pixels, which is exactly why not being able to tell by looking is not the obstacle it seems. The core moves are provenance and lateral reading. Provenance means tracing a piece of media back toward its original source: who first posted it, where, and when, rather than taking a forwarded or screenshotted version at face value. Lateral reading means leaving the image itself to see what other, reliable sources say about it, whether reputable outlets are reporting the same event, whether fact-checkers have addressed it, what a reverse-image search turns up about where it has appeared before. Students also ask who made it and why, and whether it conveniently benefits someone, and they learn to treat a single unverified item, however convincing, as a claim requiring corroboration rather than proof in itself. Emerging authentication tools like content credentials add another layer over time. None of this requires detecting the fake by eye; it establishes what is trustworthy through verification, which is the skill that actually works.

    Go deeper

    Build it into a bigger day

    Workshop
    Lateral Reading & Evaluating Sources
    Workshop
    Digital Citizenship & Media Literacy
    Workshop
    Search & Information Literacy