In a recent, highly publicized incident, the popular actress Rashmika Mandanna has become embroiled in a deepfake video controversy. This deepfake video, which has rapidly spread across social media, portrays a woman entering an elevator, but her face has been digitally manipulated to resemble Rashmika Mandanna. This occurrence has raised considerable concerns and led to calls for legal actions. Even Bollywood icon Amitabh Bachchan, who co-starred with her in the movie “Goodbye,” has voiced his apprehensions regarding the rising trend of deepfake content and has advocated for legal measures.
So, what exactly is a deepfake video?
A deepfake is a portmanteau of “deep learning” and “fake.” It refers to a video that has undergone alterations through algorithms to replace the individual featured in the original footage with someone else, often a public figure, in a manner that makes the video appear genuine. Deepfakes employ a form of artificial intelligence known as deep learning to create fabricated scenarios or events that never actually occurred. The proliferation of deepfake videos has been significantly facilitated by the accessibility of numerous AI tools, some of which are freely available and exacerbate the problem of forged photos, videos, and audio recordings.
How can one spot a deepfake video?
Although deepfake videos can be incredibly convincing, there are distinctive indicators that can aid in their identification:
- Unnatural Eye Movements: Observe for peculiar eye movements, such as a lack of blinking or erratic eye shifts.
- Color and Lighting Discrepancies: Pay attention to disparities in color and lighting between the individual’s face and the video’s background.
- Audio Quality: Compare the quality of the audio with the lip movements to check for synchronization.
- Visual Inconsistencies: Scrutinize the video for any visual inconsistencies, such as peculiar body proportions, artificial facial expressions, unusual positioning of facial features, or awkward posture and physique.
- Reverse Image Search: Perform a reverse image search on the video or the person to determine their authenticity.
- Video Metadata: Examine the video’s metadata to ascertain whether it has been tampered with or edited.
- Deepfake Detection Tools: Utilize deepfake detection tools available through online platforms or browser extensions that can flag suspicious videos.
Numerous technologies and solutions are under development to address the issue of deepfake content:
- AI-Based Detection: Many tools employ AI algorithms to identify alterations in videos. Microsoft, for example, has developed a tool that assesses photos and videos, offering a confidence score regarding their authenticity, utilizing a public dataset from Face Forensics++ and tested through the Deepfake Detection Challenge Dataset.
- Browser Plugins: Initiatives like the AI Foundation’s Reality Defender and SurfSafe browser plugins are designed to identify deepfake content online.
- Startup Innovations: Various startups are pioneering innovative solutions to combat fake content. OARO, for instance, offers tools for authenticating and verifying digital identities, compliance, and media. Sentinel is actively addressing issues related to information warfare.
- Unfakeable Records: OARO Media has introduced a system for creating immutable data trails that empower businesses, governmental bodies, and individuals to authenticate any photo or video.
The Rashmika Mandanna deepfake controversy emphasizes the immediate necessity for legal and regulatory measures to counter the spread of such deceptive content. Although these technological advancements are progressing, they are not yet universally accessible or entirely foolproof. Ultimately, the responsibility for identifying and refraining from sharing deepfake videos still rests with the users.