New Dubai AI model spots deepfakes in seven second clips
Open source deepfake detector records up to 91 per cent accuracy in tests
Dubai: The Dubai Electronic Security Center (DESC) has unveiled “SARAAB”, an artificial intelligence model designed to detect deepfake videos, marking what the centre describes as the first open source deepfake detection AI model developed by a government entity in the Arab region.
The model was officially launched on September 16, 2026, during the opening day of GISEC Global 2026 at the Dubai Exhibition Centre in Expo City Dubai.
Developed entirely by an Emirati team at the Dubai Electronic Security Center, SARAAB is part of the centre’s broader efforts to create advanced AI-driven cybersecurity solutions that can keep pace with rapidly evolving digital threats and strengthen trust in the online environment.
Get updated faster and for FREE: Download the Gulf News app now - simply click here.
Developed entirely by an Emirati team at the Dubai Electronic Security Center, SARAAB is part of the centre’s broader efforts to create advanced AI-driven cybersecurity solutions that can keep pace with rapidly evolving digital threats and strengthen trust in the online environment.
The launch comes at a time when artificial intelligence technologies are becoming increasingly capable of producing highly realistic fake videos, making it more difficult for users, organisations and platforms to distinguish between authentic and manipulated content.
Up to 91 per cent detection accuracy
According to DESC, SARAAB achieved up to 91 per cent accuracy in detecting deepfake content and recorded 89 per cent accuracy when tested on previously unseen data that was not included in its training phase.
The centre said comparative testing showed SARAAB delivered the highest accuracy among the open source deepfake detection models evaluated during its assessments.
One of the model’s most notable features is its ability to analyse and detect manipulated content in video clips as short as seven seconds.
By comparison, open source models included in the centre’s testing generally required videos of 15 seconds or longer to perform effective analysis.
Open source release planned
SARAAB can detect deepfake videos as short as seven seconds, giving it an advantage over open source models tested by the Dubai Electronic Security Center (DESC), which require longer footage for analysis.
The capability is particularly significant as short form videos spread rapidly across social media and digital platforms, making it increasingly difficult to verify whether visual content is authentic or manipulated.
DESC plans to make SARAAB publicly available through Hugging Face and GitHub, allowing AI engineers, cybersecurity professionals and researchers to test the model, build on its capabilities and contribute to its development.
The open source release is intended to foster collaboration across the global AI and cybersecurity community and support the continued development of deepfake detection technology.
The development of SARAAB marks a step forward in the UAE’s drive to build home grown cybersecurity capabilities, with the Dubai Electronic Security Center harnessing artificial intelligence to anticipate emerging threats, protect the community and strengthen the country’s digital security.
