Shakti and Google Fight Against Deepfakes

Deepfakes are fake videos or images that have been generated using deep learning techniques, such as swapping or manipulating the face of a person. They pose a serious threat to the trustworthiness of online information, especially during political events such as elections. To combat this problem, Google has partnered with Shakti, India Election Fact-Checking Collective, to detect and expose deepfakes and misinformation leading to the general elections in India.

The collaboration, active from 1 March until the end of the election period, provides a platform for fact-checkers and Indian language content publishers to share research resources, fact-checks, and alerts on viral misinformation. During the election period, Google will offer training in advanced fact-checking methods, deepfake detection, and tools like the Fact Check Explorer. The Fact Check Explorer is a tool that allows users to search for fact checks from authoritative sources on various topics, such as politics, health, and science.

The partnership also aims to amplify verified information on their platforms, in the wake of a collective agreement among major tech companies, including Google, Adobe, Microsoft, Meta and TikTok, to combat deceptive AI use, especially in deepfakes during political events. The agreement was signed in January 2024 and outlines a set of principles and best practices for responsible AI development and deployment.

Shakti is a consortium of Indian fact-checkers and news publishers, including members like the Misinformation Combat Alliance, The Quint, India Today Group, Boom, and the Press Trust of India. The consortium aims to combat misinformation in various domains, such as health, education, and environment. One of its members, the Misinformation Combat Alliance, recently collaborated with Meta to launch a dedicated fact-checking helpline on WhatsApp. The helpline is available in multiple languages and enables users to report deepfakes via a WhatsApp chatbot.

This effort aligns with a broader push by tech companies and government advisories, including measures from the Ministry of Electronics and Information Technology (MeitY), to ensure responsible AI use and prevent bias and discrimination in the electoral process. MeitY issued an advisory in February 2024 that required intermediaries and platforms to ensure that their services do not permit any bias or discrimination or threaten the integrity of the electoral process via the use of AI models.

Recent Blog : India’s Banana Fiber Eco Wound Dressings