Get Rid of Deep Fakes And Ai-generated Misinformation? (Google's Analysis)

Google collaborates with India to combat deep fakes, implementing guardrails and disclosure for AI content on YouTube, enhancing content moderation.

Written by: Krutarth Ghosalkar

Edited by: Krutarth Ghosalkar

Last Updated: December 29, 2023

Summary: Google said it would work with the Indian government to combat deep fakes and AI-generated misinformation. We also highlight existing and upcoming guidelines to help identify and remove digitally generated and modified content on the Internet.

Google, one of the world's largest technology companies, recently shed light on how it is fighting deep fakes and AI-generated misinformation in India. The tech giant acknowledged that "there is no panacea to combating deepfakes and AI-generated misinformation," but said it uses a combination of machine learning and human reviewers to combat AI-generated content and flag it.

The company is already addressing the issue of fake images with SynthID, an embedded watermarking and metadata labeling system that flags photos created using the company's text-to-image generator Imagen. We have implemented several guardrails and safeguards.

As for YouTube, creators will soon be able to disclose whether they've used modified or AI-generated content or AI tools, and let users know by adding appropriate labels in description fields and video players. must be notified. They're also working on a new "privacy request process" that will allow users to remove content if it uses AI to imitate a person's face or voice.

The tech giant also recently updated its election advertising policy to publicly disclose whether ads include content that is digitally altered or generated with the intent to deceive, mislead, or defraud users. Google Search already has several guardrails in place, including knowledge panels and featured snippets that flag deep fakes and AI-altered content.

At the upcoming Global Partnership on Artificial Intelligence (GPAI) Summit, Google said it will work with the Indian government to address security risks posed by "synthetic media" such as deep fakes and disinformation campaigns.

Additionally, the company uses a combination of human reviewers and machine learning to quickly identify and remove content that violates these policies, and leverages AI to quickly improve the accuracy of its content management systems.

Google said it is consulting with policymakers, researchers and experts across India. Last December, the tech giant provided a $1 million grant to the Indian Institute of Technology in Madras to establish the first Center for Responsible AI to examine bias in AI from an Indian perspective.

Highlighting some of its efforts, Google has also begun testing a wide range of security tools against photorealistic synthetic audio or video content generated by AI, so-called "synthetic media."

In India, vulgar deep fake videos of several celebrities including Rashmika Mandana, Katrina Kaif and Alia Bhatt were leaked on social media platforms earlier this month. Common people, celebrities and even Prime Minister Narendra Modi have expressed concerns about people's privacy and security on social media platforms.