Week 15 — Synthetic Media: Seeing Isn't Believing¶

At a glance


Date	Friday, December 4, 2026
Unit	2 — AI Unmasked
Big idea	Images, audio, and video can be generated or altered, so media needs context and verification.
Goal	Students learn practical habits for evaluating synthetic or manipulated media without becoming cynical.
Student accounts needed	None. Teacher-led examples only.
Prep time	5 minutes

Why this lesson looks the way it does¶

This lesson lands after Thanksgiving break, so it needs a clean re-entry. Do not start with panic about deepfakes. Start with the continuity: AI systems generate likely outputs, and images are another kind of output.

Students already live in visual media. The teachable goal is not "never trust pictures." That is both false and exhausting. The goal is slow down, look for context, and verify before sharing.

Use examples that are safe, non-political, and not humiliating. Synthetic media discussions can get weird fast if students start naming real classmates or public figures. Keep the examples generic: a fake animal photo, an impossible classroom, an altered product image, a voice clone scenario described but not performed.

CSTA 2026 standards¶

Concept area	Connection
Algorithms & Design	Students apply a media-check routine to evaluate visual claims.
Programming	Students connect generated media to systems that produce outputs from learned patterns.
Data & Analysis	Students distinguish image evidence from surrounding metadata, source, and corroboration.
Systems & Security	Students discuss impersonation, sharing, and account-safety implications.
Computing & Society	Students reason about trust, consent, and responsible sharing in a media-rich world.

"I can…" objectives¶

By the end of today, students can say:

I can explain what synthetic media is.
I can name clues that an image or video needs verification.
I can use context, source, and corroboration before sharing media.
I can explain why "seeing" something online is not the same as knowing it happened.

Teacher prep — the 5-minute version¶

Open a blank slide or document with the title: Seeing is evidence, not proof.
Optional: open https://commonsense.org for teacher reference on media literacy, or use your own preselected safe examples.
Write on the board:

Before break: AI can sound confidently wrong.
Today: media can look confidently real.
Write the check routine:

Pause → Source → Context → Corroborate
Prepare 2–3 safe example descriptions if you do not want to show images:
a photo of a shark on a flooded highway
a video clip with missing context
an AI-generated animal that does not exist

Minute-by-minute run sheet¶

0:00–0:04 — Anchor¶

"Two Fridays ago, before Thanksgiving break, we studied hallucinations: confident text that can be wrong. Today we apply the same habit to media. A picture can be useful evidence, but online it is not automatic proof."

0:04–0:12 — Mini-lesson: synthetic does not always mean harmful¶

Say:

"Synthetic media means images, audio, or video made or changed with digital tools, including AI. Some of it is harmless or helpful: movie effects, game art, accessibility tools, design mockups. Some of it can mislead people. The question is not 'Is technology bad?' The question is 'What claim is this media asking me to believe?'"

Write:

"Media claim: This happened, here, now, to this person."

Then:

"Those details are checkable. Where did it come from? What is the context? Does another reliable source show the same thing?"

0:12–0:36 — Paired work: media detective cards¶

Pairs analyze three scenarios. Read or project them.

Scenario A: A dramatic image shows a shark swimming down a city street after a storm. The caption says, "This happened today near us."

Scenario B: A short video shows a student knocking over a project, but it starts halfway through and has no sound.

Scenario C: A realistic image shows a new school robot cleaning classrooms. The caption says, "Eastwind is replacing janitors with robots next week."

For each, pairs answer:

What claim is the media making?
What would you check before sharing?
What is one harmless explanation?
What is one misleading explanation?

Say:

"Your job is not to be the person who says 'fake' about everything. Your job is to be the person who asks for context before spreading it."

Circulate:

"What exactly is the claim?"
"Where would evidence come from?"
"Could the media be real but the caption misleading?"
"What harm could sharing too fast cause?"

Use Scenario B for nuance.

"Can a real video still mislead? How?"

Land the point:

"Synthetic media is one issue. Missing context is another. A real clip can mislead, and a generated image can be clearly labeled and harmless. The habit is the same: pause, source, context, corroborate."

0:42–0:45 — Teaser + prompt¶

"Next Friday we switch from analyzing AI to designing with it. You will pitch an AI for Good idea on paper — no computers needed."

Tiered challenges¶

🟢 Starter (grades 5–6): Identify the claim a picture or video is making.
🟡 Builder (grades 7–8): Explain how a real image can still mislead if the caption is wrong.
🔴 Extender (grades 9–10): Compare three risks: generated media, edited media, and real media with missing context. Which is hardest to detect and why?

"What should you check before sharing media that might change what people believe?"

🤔 Think About This¶

Before you share an image or video this week, pause for two seconds.

Ask: what claim would someone believe if I shared this without context?

🔌 Unplugged Plan B — if the internet is down¶

This lesson works fully offline.

Time	Activity
0:00–0:04	Same post-break anchor.
0:04–0:12	Mini-lesson: synthetic media and the check routine.
0:12–0:28	Pairs analyze the three written media scenarios.
0:28–0:36	Groups create one misleading-caption scenario using a harmless drawing on paper.
0:36–0:42	Trade scenarios; another group identifies the claim and what to check.
0:42–0:45	Same teaser and Think About This.

Common problems & fixes¶

Problem	Fix
Students become cynical: "Everything is fake"	Say: "The goal is not disbelief. The goal is proportionate checking."
Discussion turns to embarrassing classmates	Stop it immediately: examples stay generic and respectful.
Students focus only on visual clues	Add: source, date, location, and corroboration often matter more than pixel clues.
Students ask for deepfake tools	Do not demo impersonation tools. Keep the lesson on evaluation and responsible sharing.
Political examples derail the room	Use neutral school/weather/product scenarios instead.

Vocabulary¶

Synthetic media — media created or altered with digital tools, including AI
Deepfake — synthetic media that imitates a real person's appearance or voice
Context — information around media, such as date, location, source, and purpose
Corroboration — support from another reliable source
Metadata — information about a file, such as time, location, or device, when available

Reflection — fill in after class¶

Did students stay skeptical without becoming cynical?
Which scenario produced the best reasoning?
Did any example feel too sensitive for this age span?
Are teams ready for a positive design sprint next week?

Next week: Week 16 — AI for Good: A Design Sprint