Free mini-course: five short lessons in your browser
Spot AI fakes in 5 short lessonslabels, reverse search, voice-clone scams, video tells and dated originals
TimeBinder adds a layer of protection against AI and human fraud and disputes. Five short lessons on what image labels, reverse searches and situation checks can show, and where they stop. Progress stays in this browser.
No checker on this page can reliably identify synthetic media. A TimeBind does not identify synthetic media. The lessons teach habits, not a verdict on any one file.
Start the five lessons
Each lesson is short and ends with one or two checks. Your progress is saved in this browser only. Nothing you choose is sent to TimeBinder.
Your progress is saved in this browser only. Choices stay on this device and are not sent to TimeBinder or tied to your name.
Lesson 1 of 5. 0 of 5 complete.
Lesson 1. Image labels and their limits
A Content Credentials (C2PA) manifest can record claims about how a file was made or edited, if a tool wrote one and later steps did not strip it. A SynthID-style watermark is a hidden signal used by some generators; a public checker may or may not find it, and a miss does not settle the question. EXIF and IPTC fields can name a camera, a time, or a digital-source type, and they are easy to edit or remove.
Check 1. Match each label to what it can show
Tap a label, then tap the meaning that fits it best.
Check 2. A file has no C2PA manifest
What can you conclude?
Lesson 2. Reverse image search, and what it cannot prove
A reverse image search can show earlier or matching copies, a source page, or a cropped original. That is useful when a clip is going around. A hit does not prove the first source is genuine. A miss does not prove the file is new, unique, or camera-original. Search indexes are incomplete, and new files take time to appear.
Check 1. A dramatic still is spreading today
Which next step is the most useful first move?
Check 2. The search finds no earlier copy
What does that result show?
Lesson 3. Voice-clone scam checks
You usually cannot hear a cloned voice reliably. Treat the situation: unexpected urgency, a request for money or codes, a refusal to switch to a known number, or a story that cuts you off from checking. Hang up. Call the person back on a number you already have. A family or workplace safe word helps. This lesson does not identify a cloned voice from audio.
Check 1. Tick the useful checks
Select every item that is a useful check. Leave out the ones that are not reliable on their own.
Check 2. A familiar voice asks you to send money now
What should you do first?
Lesson 4. Video tells, and why they are unreliable
Hands, teeth, blinking, typography and “too-smooth” skin were common talking points. Many current clips no longer show those artefacts, and real footage can show odd details too. A more useful check is the situation: does the same clip appear in several cities, is there a named source, does a product or event check out? Looking harder at pixels will not reliably settle the question.
Check 1. Which claim is honest?
Check 2. A viral product clip looks slightly “off”
What is the most useful next step?
Lesson 5. Keep your own dated originals
The most useful record is one you make yourself, early. Keep the original file, not only a compressed share copy. A TimeBind records that an exact file existed no later than a recorded time and that a later copy still matches its SHA-256 fingerprint, anchored to Bitcoin. It does not show ownership, that the contents are true, or whether a file is synthetic. First Pooled TimeBind is free, then from US$5. An ASAP TimeBind is US$15.
Check 1. What can a TimeBind show?
Check 2. When is the most useful time to bind a file?
- Published photo record checklist
- SHA-256 file hash checker
- How to tell if a file has been changed
- Glossary: existed no later than, SHA-256
You have finished all five lessons
Progress stays in this browser. You can revisit any lesson above. If a file matters, date the exact original while you still have it.
What does this mini-course actually teach? The short answer
It teaches five habits: read image labels (C2PA, SynthID and EXIF) without over-reading them; use reverse image search for copies and sources, and treat a miss as unsettled; check a voice-clone scam by the request, not by listening for a glitch; treat popular video “tells” as unreliable; and keep your own dated original before you only have a share copy.
It does not identify whether a file is synthetic. A TimeBind of an original you already hold can show that exact file existed no later than a recorded time and still matches. It does not show that the contents are true, or that the file is or is not synthetic.
Why these five lessons, in November 2026
Search suggestions checked on 12 November 2026 still cluster around “how to tell if an image is AI generated”, “what is C2PA”, “SynthID watermark”, “reverse image search AI”, “how to tell if a voice is AI”, “how to tell if a video is AI” and “how to prove a photo is real”. On the same day, an X search for C2PA, Content Credentials or SynthID returned a full first page of posts inside about eight hours, and first pages for reverse-image and “is this video AI” questions were similarly busy, with further pages available.
A YouTube slice of the first 20 results, filtered to uploads from 13 October to 12 November 2026, included a 5 November how-to on checking a photo (47,632 views in that pull) and several thousand views on SynthID explainers and “how to tell” guides. That slice is not a complete census: older popular videos can fill the first 20 results. The course follows those demand clusters, and it spends as much time on limits as on the checks themselves, because over-claiming is the usual failure.
This page sits under the public guide to protecting yourself from AI forgeries, next to the existing quizzes and checkers. It is not a replacement for those tools, and it is not a classroom credit course. A follow-up, Verify a viral video in 5 steps, is a workflow for checking a single clip. For a voice message or call that asks you to act, see Check a suspicious voice message in 5 steps. For a screenshot or document, see Verify a screenshot or fake document in 5 steps. For a photo of an event, see Verify an AI-generated photo of an event in 5 steps. For a clip that looks like a news report, see Verify a viral news clip in 5 steps. For a leaked-looking chat or email screenshot, see Verify a leaked-looking chat or email screenshot in 5 steps. For a product review or testimonial, see Verify a fake product review or testimonial in 5 steps. For an invoice or receipt, see Verify an AI-generated receipt or invoice in 5 steps; for a profile photo, see Check a suspicious profile photo or dating profile in 5 steps; for a rental or property listing photo, see Verify an AI-edited rental or property listing photo in 5 steps; for a celebrity or expert video ad, see Check a suspicious celebrity or expert video ad in 5 steps; for a file offered as evidence, see Check a photo, video or audio file offered as evidence in 5 steps; for a suspicious video call or remote interview, see Check a suspicious video call or remote job interview in 5 steps; for a suspicious website or login page, see Check a suspicious website or login page in 5 steps; for a map or satellite image, see Check a map or satellite image that may be AI-made or edited in 5 steps.
What each lesson can show, and what it cannot
The honest job of this course is to slow a snap judgement and to point you to records you can still show. See also synthetic media, watermark and existed no later than in the glossary.
| Lesson | What it can show | What it cannot show |
|---|---|---|
| 1. Image labels | What a C2PA / Content Credentials manifest, a SynthID-style watermark, or EXIF fields may record when they are still present. | Whether a file with no label, or a stripped label, is synthetic or genuine. |
| 2. Reverse image search | Earlier or matching copies, a source page, or a crop, when the index has them. | That a miss means the file is new, unique, or camera-original. Indexes are incomplete. |
| 3. Voice-clone scams | Useful situation checks: urgency, money or codes, a callback on a number you already have, a safe word. | Whether a voice is cloned from the sound of the call. |
| 4. Video tells | Why popular artefacts are a weak test, and why the situation around a clip is more useful. | A reliable visual rule for identifying a synthetic clip. |
| 5. Dated originals | What a TimeBind of an exact file can show: it existed no later than a recorded time, and a later copy still matches its SHA-256 fingerprint. | Who made the file, who it belongs to, that the contents are true, or whether the file is synthetic. |
If a file is one you may need to stand behind later
The most useful record is one you make yourself, early, on the exact original:
- Keep the original file, not only a compressed share copy.
- Make a TimeBind of that exact file. The fingerprint is made in your browser and the file stays on your device. You can later download a Proof of Time Certificate.
- Store the file and the certificate together, and use the published photo record checklist or the file-changed answer when you later need to compare a copy.
First Pooled TimeBind free, no card, then from US$5. An ASAP TimeBind is US$15. A TimeBind does not identify synthetic media.
First Pooled TimeBind free, no card, then from US$5. An ASAP TimeBind is US$15.
What a TimeBind shows, and what it does not. It shows that an exact file existed no later than a recorded time and that a copy shown later still matches its fingerprint. It does not show who made the file, who owns it, that its contents are true, or whether AI was used to make or edit it. This is general information, not legal advice.
Spot AI fakes in 5 short lessons: FAQ
Can this course tell me whether a file is synthetic?
No. No checker on this page, and no tool on this site, can reliably identify synthetic media. The lessons explain what labels, reverse searches and situation checks can show, and where those checks stop.
Does a TimeBind identify synthetic media?
No. A TimeBind shows that an exact file existed no later than a recorded time and that a later copy still matches its fingerprint. It does not show who made the file, who the file belongs to, that the contents are true, or whether the file is synthetic.
Why are “tells” in photos and videos unreliable?
Classic artefacts such as odd hands, blinking or scrambled text can be missing in current clips, and genuine footage can look odd too. A closer look at pixels is not a reliable test. The more useful questions are about the source and the situation around the file.
Where is my progress saved?
In this browser only, using localStorage. Choices are not sent to TimeBinder and are not tied to your name. Clearing site data, or using another browser, starts the lessons again.
Is the course free?
Yes. The five lessons and the checks run in your browser at no charge. If you later make a TimeBind, the first Pooled TimeBind is free, then from US$5. An ASAP TimeBind is US$15.
Related guides and tools
- How to protect yourself from AI forgeriesThe public AI-fakes guide this course sits beside
- Content Credentials checkerWhat a C2PA manifest can and cannot show
- AI watermark and label decoderWhat a SynthID result or a platform label can show
- AI voice scam red-flags checkerSituation checks for a suspicious call
- Viral product video checkerA situation check for a viral product clip
- Real or AI? photo quizPractice rounds; spotting by eye is unreliable
- C2PA vs SynthID vs a dated file recordWhat each kind of record shows
- GlossaryPlain definitions of C2PA, SynthID, EXIF and existed no later than
Date the exact originals you still have
Add a layer of protection against AI and human fraud and disputes: a dated, checkable record of the exact file, kept while you still have the original.
- Open the app and sign in with email or Google.
- Add your files. They stay on your computer.
- Seal them and download your Proof of Time certificates.
First Pooled TimeBind free, no card, then from US$5. See pricing.