Information on the cyber dangers generated by Deep Fake systems
What does Deep Fake mean?
The term “Deep Fake” refers to the advanced artificial intelligence technique used to create or modify video and audio content so that it appears real, although it is completely fabricated. This involves the use of machine learning algorithms and neural networks to imitate a person’s voice, facial movements and behaviour.
How severe are the consequences of Deep Fake?
The consequences of using Deep Fake technology can be extremely serious, ranging from the spread of fake news and manipulation of public opinion to fraud and harassment. They can affect personal and professional reputation, influence political elections and cause significant financial losses.
Example of Deep Fake:
A notorious example of Deep Fake involved a case from Hong Kong, where a person from the finance department was tricked into paying 25 million dollars after receiving a Deep Fake video. In this video, an authority figure appeared to be asking him to transfer the funds, misleading him through the apparent realism of the images and sound.
How can you tell that you are being fooled by a Deep Fake?
To identify a Deep Fake, it is important to pay attention to the fine details of the video or audio recording:
-
Check the synchronisation of lip movements with the audio.
-
Check for abnormal movement of the fingers.
-
Observe any anomalies in facial expressions or eye movements.
-
Listen for any distortions or anomalies in the voice.
-
Look for inconsistencies in the video background.
Tips to protect yourself from a Deep Fake:
-
Always check the source of any dubious material.
-
Use Deep Fake detection software, which is becoming increasingly available.
-
Do not share suspicious videos or audio without verification.
-
Educate yourself and others about the risks associated with Deep Fakes.
-
establish a Master Password which will be used to verify the authenticity of instructions given by phone or during video conferences.
What to do if you have been a victim of a Deep Fake:
-
Report the incident to the competent authorities.
-
Inform the platforms on which the material was distributed so that it can be removed.
-
Consult a lawyer to assess the legal options for protecting your image and recovering any damages.
EU legislation on protection against Deep Fake:
The European Union is intensifying its efforts to combat the spread of Deep Fakes through various directives and regulations, such as GDPR, which protects personal data and anonymity, and new legislation aimed at combating online disinformation and protecting citizens against manipulated media content.
Frequently asked questions
What is a deep fake?
An advanced AI technique (machine learning and neural networks) that creates or modifies video and audio so it appears real although it is fabricated — imitating a person's voice, facial movements and behaviour.
How severe are the consequences?
From disinformation and manipulation of public opinion to fraud and harassment. A notorious Hong Kong case: a finance-department employee was tricked into paying 25 million dollars after a deep-fake video call in which an authority figure appeared to request the transfer.
How can I tell it is a deep fake?
Check lip-sync against the audio, abnormal finger movements, anomalies in facial expressions or eye movement, distortions in the voice, and inconsistencies in the background. And always verify the source of any dubious material.
How do I protect my organisation?
Verify the source of any dubious material, don't share suspicious content unverified, train staff, and — key for fraud — establish a "master password" to verify the authenticity of instructions given by phone or in video calls. If you were a victim, report it to the authorities and platforms.