Kaspersky's Global Research & Analysis Team (GReAT) has just discovered ads that provide real-time deepfake video and audio services on darknet. The starting price is recorded as about 50 USD (about 1.3 million VND) for fake videos and 30 USD (about 800,000 VND) for fake voice messages, and can be higher depending on the complexity and duration of the content.
Previously, deepfake creation services on darknet ranged from $300 to $20,000/minute. Deepfake packages now allow bad guys to quickly create fake audio and videos in real time at a very small cost compared to before.
The ads also promise many features, including real-time face swapping during a video call on an online or messenger platform, replacement of the face for authentication, as well as forging the source of the camera on the device.
Some advertising posts claim to be able to provide software to synchronize a person's facial features in videos with dialogue, even in foreign languages. They also sell voice cloning tools, allowing voice impersonation and adjusting loudness and tone to express different emotions.
However, it is likely that these advertisements are actually just a scam to appropriate money from people who intend to buy.
Mr. Dmitry Galov, Head of Kaspersky Global Research & Analysis in Russia and CIS, commented: We not only see advertisements providing deepfake services, but also see the increasing demand for these tools.
The suspects are actively experimenting and integrating AI into their activities. Some platforms have even developed more sophisticated features, such as a completely built-in malicious big language model, separate from large, open-ended language models that can run directly on criminal devices.
According to Mr. Galov, although these technologies have not created a completely new cyber threat, they make attackers significantly more dangerous. In that context, cybersecurity experts need to make further efforts to respond. One of the most promising directions is to take advantage of AI to both increase the productivity of security experts and improve defense efficiency.
To protect safety against deepfake threats, Kaspersky recommends:
- Enterprises fully deploy cybersecurity measures: not only stopping at installing protection solutions, but also need a team of highly skilled IT experts.
- Ensure employees understand deepfake and the dangers of this technology, and regularly organize training courses to help employees know how to identify it.
- Equip staff with skills to recognize typical signs of deepfake: adjusted movements, not smooth, inconsistent lighting between frames, unnatural skin tones, unusual blinking or little blinking, distorted images, videos intentionally filmed at low quality or poor light.
