IT Glossary
What is a deepfake?
A deepfake is audio, video or imagery fabricated with artificial intelligence, in which a real person appears to say or do something that never happened.
A finance manager takes a video call. The chief executive is on screen, the voice is right, the manner is right, the request is urgent and confidential: move the funds today, the acquisition closes tomorrow. Everyone on that call was synthetic except the manager. Losses of this kind are no longer hypothetical; they have been reported in the tens of millions. A deepfake is media generated or altered by machine learning so that a real person's face, voice or handwriting is convincingly reproduced. Three minutes of a conference talk online is enough source material for a usable voice clone, and a public photo set is enough for a face. What changed is not the existence of forgery but its price, which has fallen to roughly nothing. That is why this belongs in an operations manual rather than a technology newsletter: detection by eye is unreliable and getting worse, so the defence that survives contact with reality is procedural. Payment approvals requiring a second channel, callbacks to numbers from your own records rather than the ones offered, and a stated rule that no genuine executive will ever be offended by verification.
Let’s talk about your project
Message us on WhatsApp or send an email — you talk directly to a developer.
office@northdan.com · +40 752 070 247
Why it matters for your business
Fraud stopped by process, not luck
Two-channel confirmation on payments and supplier changes defeats the attack regardless of how convincing the fake looked or sounded.
Staff who are hard to rush
People who have seen a convincing forgery in training recognise the pressure pattern, and pressure is the part attackers cannot remove.
Reputation defended in advance
A pre-agreed response — who verifies, who speaks, which channels are official — turns a fabricated clip from a crisis into an incident.
Frequently asked questions
How do I recognise a deepfake video or voice call?
Visual tells are fading, though some persist: mismatched lighting at the jaw and hairline, blinking that looks scheduled, teeth and jewellery that shimmer, lip movement drifting on hard consonants. Audio fakes still struggle with laughter, interruption and background noise. The reliable test is behavioural: ask something only the real person would know, or hang up and call back on a stored number.
What do we do if our company is targeted by deepfake fraud?
Freeze the payment path first and contact your bank immediately, because recall is sometimes possible within hours. Preserve the recording, call metadata and message headers; file with the police cybercrime unit and, if personal data was involved, assess whether the supervisory authority must be notified. Then tell staff plainly what happened, because silence guarantees the second attempt succeeds.
Are deepfakes illegal in the European Union?
The use is regulated, not the technique. Fraud, identity theft, defamation and non-consensual intimate imagery are already criminal across EU member states, and the EU AI Act adds transparency duties requiring synthetic media to be labelled. For a company the practical exposure is usually civil and reputational: your brand appearing in content you never made.
Let’s talk about your project
Message us on WhatsApp or send an email — you talk directly to a developer.
office@northdan.com · +40 752 070 247