Shallow fakes

Source– This post on Shallow fakes is based on the article “More than deepfakes, shallow fakes should worry everyone” published in “The Indian Express” on 29th March 2024.

Why in the News?

With Lok Sabha elections starting on April 19, social media is flooded with misinformation, mostly in shallow fakes forms.

What are Shallow fakes?

1. About– Shallowfakes, similar to deepfakes, involves altering people’s images, videos, and voice clips for malicious purposes. However, unlike deepfakes, which requires advanced AI software, shallowfakes can be made using basic editing software like Photoshop.

2. Why term “shallow”? – Shallow Fakes are called so because they involve lower-quality image and video editing compared to the sophisticated techniques used in Deep Fakes.

3. Used for:
a. To fabricate false proof of identity or address, such as photo ID documents like passports and driver’s licenses.
b. To produce counterfeit supporting evidence for claims or transactions, such as contracts, agreements, invoices for services, no claims discount certificates, etc.
c. Spreading political narratives and defaming opponents, for example, a video depicting AIMIM leader Asaduddin Owaisi singing Shiv Tandav stotra.

4. Impact– Due to their simpler creation process, many experts see shallowfakes as posing greater threats than deepfakes.

What are Deepfakes?

Deepfakes are fake content in the form of videos, pictures and audio which are created using powerful artificial intelligence tools. The origin of the word “deepfake” can be traced back to 2017 when a Reddit user, with the username “deepfakes”, posted explicit videos of celebrities.

Difference between deep fakes and shallow fakes

1. Technology used– Deepfakes use advanced AI and machine learning algorithms, while shallowfakes are less advanced, involving basic editing techniques like cropping, splicing, or adjusting video speed. Shallowfakes can be created with common editing software and may not require AI or machine learning.

2. Accessibility– Shallow fakes are easier for regular people to make and are used more widely compared to deepfakes, which need AI and machine learning.

3. Detection– Deepfakes are typically difficult to detect with the naked eye and may require specialized software or expertise. Shallowfakes are usually easier to spot than deepfakes because they can show visual mistakes or differences.

UPSC Syllabus- Science & Technology