A digital platform that helps you tell whether an image, audio clip or video is real or manipulated, and shows you the result clearly and quickly through digital forensics and AI.
Upload an image, an audio clip or a video. Takkad detects its type automatically and shows you the result within seconds.
Checking opens soon. This is a preview of the experience.
Be careful
92%
Confidence
Preliminary triage, not final legal evidence
If you got a call in the voice of someone you know, asking you to transfer money fast… how would you know it is really them?
What if you had three seconds to check before transferring?
Total financial fraud losses worldwide in 2025 alone
Source: INTERPOL Global Financial Fraud Assessment (March 2026)
of Saudis were targeted by a financial fraud attempt by phone or electronic means
Source: National Center for Public Opinion Polls survey (2022)
of them actually lost money
Source: National Center for Public Opinion Polls survey (2022)
only of victims report the incident
Source: Source being documented
Three steps only.
Upload the image, audio clip or video directly, or share it to Takkad from your messaging app on Android. Before any processing, the file fingerprint and the time are recorded.
Analysis starts right away. Faces, voice and lip sync are examined with the technologies we built.
The result comes with an explanation of the reasons and a confidence percentage, and a certificate with a reference number that carries the digital fingerprint and can be shared. If something looks suspicious you get immediate advice, such as not transferring money and contacting the person directly. It is a preliminary triage, not final legal evidence.
Three tracks, each with signals that Takkad shows in your result.
We look at faces and at traces of editing and compositing, and read the file metadata when it exists.
We turn the sound into a spectrogram and look for the fingerprint of generated voices, including WhatsApp voice notes.
We examine the face and the voice together, compare lip movement with speech, and follow how consistent the face is across frames.
Four technologies work together, and the result is a preliminary triage that AI explains in plain language.
Examines faces in images and video frames for traces of compositing and manipulation.
Turns the sound into a spectrogram and looks in it for the fingerprint of generated voices.
Tracks the face and lips and compares their movement with the speech.
Explains the result in plain language and suggests your next step. It explains and does not decide, and it never receives personal data.
Immediate protection before a rushed financial or emotional decision under pressure.
Complete ease of use, with no technical expertise needed.
Less reliance on instinct alone in a moment of doubt.
Supports SDAIA’s national direction against deepfakes. It is an independent project, not affiliated with it.
For organizations that handle suspicious material, Takkad provides documentation you can refer back to.
A report in Arabic and English carrying the digital fingerprint, the timestamp and a digital signature that can be verified.
Every step from receiving the file to the report is recorded with its time.
Append-only, cannot be modified.
Coming soon, to connect checks to your organization’s systems.
We collect as little data as we can, and tell you plainly what happens to your file.
Your file is deleted automatically after 24 hours and never used for training without your consent.
The digital fingerprint (SHA-256) and the time are recorded before any processing, to document that the file did not change.
Forensic reports are digitally signed, and any change to them shows up when verified.
The result is a preliminary triage, not final legal evidence. We recommend referring the file to the lab when needed.
Our plan for the year after launch. It may change as we learn.
Months 1 to 3
Months 4 to 6
Months 7 to 9
Months 10 to 12
A team of five combining project management, design, data analysis, AI and software development.
Digital projects manager, team lead
UX and UI design expert, team member
Data analytics engineer, team member
AI engineer, team member
Software development engineer, team member
We do not keep your file longer than needed. It is deleted automatically after 24 hours and never used to train models without your explicit consent.
No. The result is a preliminary triage that helps you decide. If you need an official report, the file can be referred to a forensic examiner.
This result appears when there are suspicious signals, and also when the file is not enough to settle the question, for example when its quality is low or the clip is very short. We show you the signals and advise you to check the source before acting.
Images, audio and video in common formats, including WhatsApp voice notes.
A visitor without an account can run up to 3 checks a day, and the limit rises with a free account.
Do not make the transfer or follow the request. Contact the person directly on a number you already know, then report it to the relevant authority.
Yes, it is free for individuals for now.
Use the contact link at the bottom of the page.
Upload the suspicious image, audio clip or video, and decide with the facts in front of you.
