Pak-linked profiles, AI deepfakes: Cyber probe uncovers CJP protest disinformation

A forensic probe into online activity surrounding the CJP protests uncovered a cross-border disinformation network using AI-generated videos, cloned voices and other manipulated content, with more than 500 social media profiles and 429 posts under the scanner.The accounts were spreading false and misleading narratives across Instagram, X, Facebook and YouTube with the objective of influencing public perception and disturbing public order, officials said. (Photo PTI)The Maharashtra Cyber Department has identified more than 500 social media profiles and 429 objectionable posts linked to what it described as a coordinated cross-border disinformation campaign during the nationwide Cockroach Janata Party (CJP) protests, according to an official statement issued on Monday.The state cyber agency said the network was uncovered during a technology-assisted monitoring exercise that analysed social media activity using open-source intelligence (OSINT), metadata examination, behavioural analysis and forensic tools to identify AI-generated and manipulated content circulating online.According to Maharashtra Cyber, the accounts were spreading false and misleading narratives across Instagram, X, Facebook and YouTube with the objective of influencing public perception and disturbing public order.The agency said its technical assessment indicated that around 100 of the identified accounts were being operated from outside India, particularly from Pakistan and certain countries in the Middle East. Officials said the monitoring drive led to the identification of 429 objectionable posts and social media handles, while more than 500 profiles were flagged as being part of the broader network. According to the statement issued by the department, legal action has been initiated against the identified accounts, and requests have been sent to social media platforms to remove the content.The investigation also focused on the use of artificial intelligence in misinformation campaigns. Maharashtra Cyber said over 140 posts and profiles contained AI-generated or AI-assisted content, including digitally inflated crowd visuals, synthetic voices, cloned audio and manipulated videos attributed to public figures.As an example, the department said its forensic examination of one video purportedly featuring a Union Minister detected synthetic or cloned speech in the audio stream with 100 per cent confidence, while analysis of the video indicated approximately 82 per cent confidence of facial manipulation, with nine of the 11 analysed segments showing signs of reenactment.The cyber agency said it would continue monitoring digital platforms during the protests and take legal action against those involved in spreading misinformation and deepfake content. It also urged users to verify information through official sources and credible news organisations before sharing content related to the protests.CJP ended its 36-day protest at Delhi's Jantar Mantar on Saturday after the Centre agreed to its principal demands, including the resignation of then Union Education Minister Dharmendra Pradhan, withdrawal of FIRs, protection against retaliatory action and a promise to examine wider reforms in the examination system. The government also agreed to provide a written draft of the understanding by Tuesday following legal consultations.- EndsPublished By: Shipra ParasharPublished On: Jul 28, 2026 08:24 IST