Gandhinagar – : Infibeam Avenues Ltd, a leading fintech company, has announced signing of a strategic MoU for Research and Development (R&D) with the Indian Institute of Sciences (IISc) Bangalore, one of the world’s top-ranking universities known for its research excellence. This collaboration aims to research and develop advanced real-time deepfake detection systems designed to enhance digital security for government entities, corporations, and organizations, effectively combating the rising threat of AI-generated deception.
As deepfake technology continues to evolve, it poses significant risks to personal and corporate integrity. These sophisticated AI-generated media can create hyper-realistic but false representations of individuals and events, leading to misinformation, cyberbullying, harassment, financial fraud, and identity theft. Such threats have far-reaching implications for businesses, government institutions, and the general public.
A notable incident highlighting the urgency of this issue occurred in January 2024, when a Hong Kong-based multinational company lost approximately $25 million (around Rs 207 crore) due to a deepfake scam. Scammers executed a convincing live video call using deepfake technology to impersonate company executives, leading to a severe financial loss before the deception was uncovered. (Source: South China Morning Post and Business Insider)
“Digital communications and a digital India will thrive only as long as there is trust,” said Mr. Rajesh Kumar SA, CEO of Phronetic.AI, an AI unit of Infibeam Avenues Ltd. “This partnership is a pivotal step in restoring trust in digital communications. Together, we will equip users with the necessary tools to differentiate between truth and fabrication in an increasingly complex digital landscape, thereby mitigating fraud risks and enhancing digital trust.”
Under the terms of the Memorandum of Understanding (MoU), Infibeam Avenues Ltd’s AI business unit, Phronetic.AI, and the IISc team will collaboratively develop anti-deepfake technology specifically tailored for real-time video communication. The partnership will focus on selecting the most effective detection models for various scenarios, ensuring that real-time deepfake detection operates efficiently and cost-effectively at scale.
“We are committed to staying ahead of malicious actors by developing innovative AI solutions that ensure digital authenticity,” said Mr. Vishal Mehta, Chairman and Managing Director of Infibeam Avenues Ltd. “This partnership is a crucial step toward enhancing cybersecurity and preventing the misuse of deepfake technology for fraudulent activities.”
Despite the availability of various deepfake detection tools in the market, only a limited number possess the capability for real-time operation. In a pioneering effort, Infibeam’s Phronetic.AI team has developed an advanced video AI agent that can detect deepfakes in real-time through a novel interventional technique. This agent actively engages in live video conversations, alerting users when the other participant is identified as a deepfake. Infibeam has already filed a patent for its innovative real-time deepfake detection algorithm.
Recognizing the increasing sophistication of deepfakes and the necessity for detection algorithms to evolve continuously to address this growing threat, the company has collaborated with Vision and AI Lab (VAL) of the Indian Institute of Science (IISc), where it aims to improve these algorithms further, ensuring robust defenses against the challenges posed by increasingly realistic deepfake technology.
“As Generative AI continues to advance at an unprecedented pace, the rise of deepfakes poses a significant challenge. Without proactive measures, the spread of AI-generated misinformation could become a major concern. Addressing this requires ongoing efforts from AI researchers to monitor emerging generative models and develop robust techniques to detect deepfakes effectively.” said Prof. Venkatesh Babu, Professor and Chair of the Dept. of Computational and Data Sciences (CDS) IISc.
Additionally, the research will prioritize the development of a user-friendly interface, enabling easy access for non-experts to verify the authenticity of live visuals and audio. This scalable detection system will be adaptable across various sectors, including banking, healthcare, insurance, finance, fintech, HR recruitment, government organizations, police, armed forces, and personal communications, addressing the diverse needs of industries particularly vulnerable to deepfake technology.
This research initiative aims to offer Real-Time Deepfake detection AI Agent that enhances public confidence and protect the reputations of its users whether it’s a government institutions, organizations, or a corporations.
