UAE researcher develops AI platform Deepy to detect deepfakes in images, videos and audio

Emirati cybersecurity researcher Bashayer Al Salami has launched Deepy, an AI-powered platform that detects manipulated images, videos and audio files, as deepfakes grow more sophisticated and harder for ordinary users to identify, Khaleej Times reported.

Al Salami developed the platform following postgraduate research at Zayed University into emerging cyber threats and the misuse of generative AI. The launch comes as concerns over AI-driven misinformation rise globally.

What is Deepy and how does it detect deepfakes?

Deepy is an AI platform that allows users to upload digital files and receive a probability assessment of whether the content has been altered using AI tools. It analyses technical characteristics within images, video and audio to detect synthetic manipulation. Its detection models were trained on thousands of examples of both authentic and fabricated media.

The platform examines indicators including irregular pixel structures, inconsistencies in lighting and shadows, unnatural facial expressions, and distortions in motion or sound patterns. It also identifies digital traces commonly left behind by AI-generation software. This combination of signals allows Deepy to flag manipulated content at an early stage, according to Al Salami.

"With the rapid evolution of artificial intelligence, deepfakes have become one of the most dangerous digital threats because they can create highly convincing fabricated content that ordinary users struggle to recognize," Al Salami said.

Who is Deepy designed for?

The platform is primarily aimed at government institutions, security agencies, media organizations and companies that rely on digital verification. Al Salami said broader public-access versions are also being developed so everyday users can verify online content before sharing it. Deepy seeks to give journalists, officials and social media users a rapid verification tool before manipulated content spreads widely.

Al Salami noted that periods of political unrest or crisis often trigger a surge in misleading or fabricated material online, with manipulated content frequently used to influence public opinion or fuel confusion.

Why does AI-powered deepfake detection matter now?

Deepfake technology has outpaced the ability of typical social media users to spot altered content, and law enforcement agencies across the UAE and GCC have previously warned about AI-generated deepfakes being used for scams, financial fraud and the spread of unverified geopolitical content.

Al Salami said technology must do more than flag manipulation. "It should also contribute to strengthening digital awareness and creating a safer and more trustworthy media environment," she said.