Free mini-course: five short steps in your browser
Check a map or satellite image that may be AI-made or edited in 5 stepsthe claim and the image, the provider original, other sources, clues both ways, and a dated record
TimeBinder adds a layer of protection against AI and human fraud and disputes. A five-step routine for a map or satellite image that may be generated or edited. Guidance only. Progress stays in this browser.
No checker on this page can reliably tell AI from real. A TimeBind does not show that an image is genuine or fake. The steps are a routine, not a verdict on any one image.
Start the five steps
Each step is short and ends with one or two checks. Your progress is saved in this browser only. Nothing you choose is sent to TimeBinder. The course uses invented examples only, no real place, person or event, and own-made diagrams.
Your progress is saved in this browser only. Choices stay on this device and are not sent to TimeBinder or tied to your name.
Step 1 of 5. 0 of 5 complete.
Step 1. Name the claim, the image and the path
Maps and satellite or aerial pictures are shared with claims about damage, construction, crowds, floods, fires, borders or a place that is supposedly changing. Before you judge the picture, write down three things. The claim is what the image is said to show: which place, what event and what date. The image is the exact item you hold: its format, whether it is a screenshot, a crop or an export, and what the labels, scale and north arrow say. The path is who made it, who posted it and through which accounts or apps it reached you. Keep the three apart. An image can have a real base with an invented overlay, so a familiar view of a real place does not support the claim that sits on top of it. A BBC Verify report on a retracted AI feature in a mapping app quoted an expert saying that a forgery on such a base inherits the credibility of the map. A convincing image with an unknown path is still unverified.
Check 1. Match each part to what it means
Tap a part, then tap the meaning that fits it best.
Check 2. An invented account posts an aerial view of a town with a caption that a bridge has collapsed
What is the best first move?
Step 2. Find the provider original and the details
Ask where the view came from: the imagery provider or app, the place and coordinates, the zoom level and the capture date it shows. A link to the live view or the provider’s own archive is worth more than a screenshot, because screenshots, crops and re-saves lose details. Al Jazeera Media Institute’s guide to satellite imagery advises that when you doubt an image you should verify that it matches the original source of the imaging provider, and it names free and archive sources such as the European Space Agency and the World Imagery Wayback tool, which keeps earlier versions. News wire services often publish provider imagery too. Check what date the provider’s image carries and whether it fits the date in the claim. If the sender cannot say where the view came from, or refuses, that is information, not proof. Remember that a mapping app can also create new images: BBC Verify reported that a feature added to Google Earth in July 2026 let users generate invented scenes on top of real satellite, aerial or 3D imagery that could be downloaded and shared, and that Google paused it in under 48 hours.
Check 1. Which request is most useful?
Check 2. The sender only has a screenshot and no place, date or provider
What does that change?
Step 3. Compare with other imagery and outside sources
The strongest checks sit outside the picture. Find other imagery of the same place from a different provider or date, including the archive tools above, and see whether fixed features such as roads, coastlines, buildings, field patterns and bridges match, and whether what the claim says has changed is shown there too. Compare with street-level photos, maps and ground photos that other people can confirm, and with news reports, official notices and people who were there. Use the geolocation habit of anchoring on fixed visual features, and see the glossary on image geolocation. The Al Jazeera guide also suggests checking the direction of shadows and using a sun-position tool such as suncalc.org to test whether a claimed date and time fit, and consulting a remote sensing expert. If nothing outside the picture supports it, say that it is unsupported rather than that it is fake. Maps made by an image generator rather than a mapping service are a separate case: check every place name, border and label against an authoritative map.
Check 1. Which of these are sound checks outside the picture?
Tick all that apply.
Check 2. No other source shows the change, but the image looks entirely natural
What is the honest description?
Step 4. Clues, both ways: edits and false “it is AI” claims
Some clues justify a closer look: shadows that do not fit the claimed time, scale or north that do not match, features that no other imagery of the place shows, repeated or smeared textures, labels that do not match an authoritative map, or an edge where a new scene meets the base. Provider watermarks can help: Google said that images made with its tools carry invisible watermarks and that its Gemini app or Lens can be used to check them. BBC Verify found this generally worked for the retracted feature, but also found it was possible to trick the chatbot into saying a fake image was real, and found that external AI-detection tools sometimes failed to flag the images. So a positive watermark result is a clue, a missing one means little, and a detector score settles nothing. The reverse problem is real too: genuine satellite images are sometimes dismissed as “AI” without any basis, and a bare claim is not support either. Fast-moving events make doubt more costly. Say what you found, what you could not find and what would change your mind.
Check 1. A provider watermark check on the image finds no mark
Check 2. Someone replies only “that is AI” and offers nothing else
Step 5. Document what you have and decide next steps
Honest outcomes include “matches the provider’s original and independent sources”, “unsupported, so I did not share or rely on it” and “disputed, passed to an expert”. Write down the claim, who sent the image and when, the provider and place, what you compared and what stayed open. Keep the exact file you received and work on copies, and do not re-save the only copy. Do not repost an image you could not support, even to ask whether it is fake, because the post can spread without your caveat. For anything important, such as a published report, a dispute, an insurance matter or a safety decision, involve an experienced geospatial or remote-sensing professional. In the EU, the AI Act’s transparency rules for AI-generated or manipulated content (Article 50) are stated by the cited text to apply from 2 August 2026, but a label shows only what its provider or user disclosed, and an unlabelled image is not thereby genuine. 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 that an image is genuine or fake, who made it, or that a claim is true. First Pooled TimeBind is free, then from US$5. An ASAP TimeBind is US$15.
Check 1. What can a TimeBind show about an image you were sent?
Check 2. You could not support the image. What now?
You have finished all five steps
Progress stays in this browser. You can revisit any step above. For anything important, such as a news report, a dispute or a safety decision, involve an experienced geospatial or remote-sensing professional and do not share an image you cannot support.
How do you check a map or satellite image that may be AI-made? The short answer
Work through five steps: write down the claim (what the image is said to show), the image (the exact item you hold) and the path (who made it and posted it); ask for the provider or app, the place, zoom and capture date and look for the provider’s original or archive instead of a screenshot; compare with other imagery of the same place, fixed features, ground photos, news and people on the ground, and test shadows against the claimed date and time; treat mismatches as reasons to look closer, treat watermark results and detector scores as clues only, and treat a bare claim of “AI” as unsupported too; then record what you received, asked and compared, keep the exact file, and do not share what you cannot support.
None of this tells you for certain whether an image was generated, edited or genuine. A TimeBind of an exact file can show that it existed no later than a recorded time and still matches. It does not show that an image is genuine or fake, who made it, or that its claim is true.
Why this course, in November 2026
Search suggestions checked on 13 November 2026 show interest: “ai generated map” returned 10 suggestions (including “ai generated map in school” and “ai generated map given to students”), “fake satellite image” returned 5. Suggestions show what people type, not volumes. The topic also had a recent news event. On 31 July 2026 BBC Verify reported that Google had paused a new AI feature in Google Earth, less than 48 hours after launch, that let users generate invented scenes on top of real satellite, aerial or 3D imagery. Google said it had seen screenshots that appear to violate its policies and would work on stronger guardrails.
Sources: BBC News, “Google withdraws new Earth AI tool after warnings over misinformation risks” (31 July 2026), The Guardian, TechScape, “Google Earth’s potential disinformation nightmare” (3 August 2026), Al Jazeera Media Institute, “How to analyse satellite imagery” (25 May 2023) and EU Artificial Intelligence Act, Article 50 (transparency obligations). The last is an unofficial text of the Act. Check the official sources for your matter.
This is the sixteenth course in the series. The first, Spot AI fakes in 5 short lessons, is a general introduction, Verify an AI-generated photo of an event covers event photos, and Verify a viral video in 5 steps is a workflow for any single clip. This one is narrower: a map or a satellite or aerial view, where the key moves are the provider original and comparison with other imagery. It sits under the public guide to protecting yourself from AI forgeries.
What each step can show, and what it cannot
The honest job of this routine is to slow a decision and leave you with notes and questions that an expert can use. It is general guidance, not professional geospatial advice. See also synthetic media and existed no later than in the glossary.
| Step | What it can show | What it cannot show |
|---|---|---|
| 1. Claim, image and path | What is claimed, what exact item you hold and how it reached you. | That the image is genuine, or that its claim is true. |
| 2. Provider original | Whether the provider, place, date and an original or archive view can be produced, and what a screenshot has lost. | That an original is genuine, or that a screenshot is fake. |
| 3. Other sources | Whether other imagery, fixed features, ground photos, news and people on the ground agree, and whether shadows fit the claimed time. | That an image is fake because one source is missing, or real because one agrees. Each is one input. |
| 4. Clues, both ways | What mismatches and edges suggest. A watermark result is one clue, and a watermark can be absent or defeated. | Truth or authorship. A natural look, a detector score or a bare “it is AI” settles nothing. |
| 5. Document and decide | What you received, asked, compared and left open, and that an exact saved file existed no later than a recorded time, with a later copy still matching its SHA-256 fingerprint. | That an image is genuine or fake, who made it, or that a claim is true. |
If an image may matter later
If an image could become part of a report, a dispute or an insurance matter, a dated record of the exact files you hold helps:
- Keep exactly what you received, with the post or message it came with, and work on copies. Add notes on who sent it, when, what you asked for and what you compared.
- Make a TimeBind of each 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 files, the certificates and your notes together. Use the file date and metadata inspector or the file-changed answer to compare a copy later.
First Pooled TimeBind free, no card, then from US$5. An ASAP TimeBind is US$15. A TimeBind does not show that an image is genuine or fake.
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.
Check a map or satellite image that may be AI-made: FAQ
Can this course tell me whether a satellite image is AI-generated?
No. No checker on this page, and no tool on this site, can reliably tell AI-generated or AI-edited images from genuine ones. The steps help you find the provider original, compare other sources and weigh clues both ways. “Unsupported, so I did not share it” is a valid result.
Does an AI-detector score or a watermark check settle it?
No. A watermark found is a clue, and one not found means little. BBC Verify reported that provider checks generally worked on one retracted feature but could be tricked, and that external detection tools sometimes failed. Treat scores as clues only.
Can a real satellite base carry an invented scene?
Yes. BBC Verify reported that a feature briefly added to Google Earth let users put invented scenes on top of real imagery, which could be downloaded and shared. That is why a familiar view of a real place does not support the claim laid over it.
What about maps drawn by an image generator?
Treat them as illustrations, not sources. Check every place name, border and label against an authoritative map, and do not rely on a generated map to find a route, a boundary or a location.
Does a TimeBind show that an image is real?
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 image, that it is genuine or fake, or that its claim is true.
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 steps again.
Is the course free?
Yes. The five steps 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
- Verify an AI-generated photo of an eventWhen the picture is a ground-level photo
- Verify a viral video in 5 stepsA workflow for any single clip
- Verify a viral news clip in 5 stepsWhen the claim comes with a news-style clip
- Verify a screenshot or fake documentWhen the item is an image of a page or a chat
- Spot AI fakes in 5 short lessonsThe first mini-course: labels, reverse image search, voice scams and more
- Real or AI? quizPractise on own-made and public-domain images
- File date and metadata inspectorSee what a file’s own details say
Date the exact files you keep
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 it.
- 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.