Advancements in AI are growing leaps and bounds by the day. Normally, we’d be chuffed at these developments. Unfortunately, as useful as AI is, we’ve seen it used for malicious and nefarious purposes. Deepfakes are an example of AI used for the wrong reasons. As AI gets better, so do deepfakes. Thankfully, HONOR wants to fight back against this by announcing that its AI Deepfake Detection technology will be rolling out globally this coming April.
HONOR deepfake detection
If you’re hearing about this for the first time, HONOR debuted its AI Deepfake Detection technology back in IFA 2024. According to HONOR, its proprietary technology will be able to detect pixel-level synthetic imperfections, border compositing artifacts, inter-frame continuity issues, and facial anomalies such as face-to-ear ratio, hairstyle, and facial features.
When HONOR’s system detects that there is deepfaked material, users will receive an immediate warning. This will allow them to make more informed decisions and avoid getting scammed. For instance, if you’re watching a video you know is a deepfake but done for comedic reasons, you can choose to ignore this warning. However, if there is a video circulating claiming to be a politician or celebrity making some announcement, then paying attention to HONOR’s warning system could come in handy.
Why this matters
As we said, as useful as AI is, the potential for misuse is there. HONOR found some disturbing deepfake trends. For instance, from November 2023 to November 2024, 49% of companies encountered deepfake incidents. Also, a staggering 59% of people struggle to differentiate between content that is real, and those generated by AI. 61% of executives have also admitted that they haven’t established any protocols that address the risks of deepfakes.
The company also found the year-on-year growth of deepfakes from Q1 2023 to Q1 2024 is staggering. It saw a 900% increase in marketplaces, 533% in fintech, and 1,520% in iGaming.
According to UNIDO Representative Marco Kamiya, “The use of AI Deepfake Detection on mobile devices is a powerful weapon to fight against the deepfake, which can detect flaws in deepfake content that might go unnoticed by the human eye, such as eye contact, lighting, image clarity, and video playback, deterring people from continued engagement with the potential risk for their safety. This allows people, SMEs, and industries, to enjoy the convenience of digital communication without worrying about privacy leaks and fraud risks.”
