#FactCheck - Viral Video Falsely Claims RSS Chief Mohan Bhagwat Called for ‘Saffronisation’ of Indian Army
A video purportedly showing Rashtriya Swayamsevak Sangh (RSS) chief Mohan Bhagwat making remarks about the “saffronisation” of the Indian Army has been widely circulated on social media. The clip claims that Bhagwat called for the removal of non-Hindus from the armed forces and linked the issue to future political leadership changes in the country.
Claim
However, a verification by the Cyber Peace Foundation has established that the video is misleading and has been digitally manipulated.
In the video, Bhagwat is allegedly heard saying that unless more than 50 percent of non-Hindus are removed from the Indian Army by 2028, Prime Minister Narendra Modi would be replaced by Uttar Pradesh Chief Minister Yogi Adityanath. The clip further attributes another statement to him, suggesting that he would resign if the Prime Minister were to demand Nitish Kumar’s resignation.
By the time of publication, the video had been viewed over 7,000 times.( lINK, ARCHIVE Link, Screenshot
Fact Check:
The reverse image search also directed the Desk to a video uploaded on CNN-News18’s official YouTube channel on December 21, 2025. The footage was found to be a longer version of the viral clip and was recorded at the RSS centenary event held in Kolkata on the same date. A comparison of both videos confirmed that the background visuals, stage setup and camera angles were identical.
However, a careful review of the original CNN-News18 video revealed that Mohan Bhagwat did not make any of the statements attributed to him in the viral clip.
In his original address, Bhagwat spoke about unity and referred to concerns over increasing atrocities against Hindus in Bangladesh. He made no reference to the Indian Army, nor did he comment on its composition or alleged saffronisation. Here is the link to the original video, along with a screenshot: https://www.youtube.com/watch?v=KnsAUGfBQBk&t=1s
In the next phase of the investigation, the audio track from the viral video was extracted and analysed using the AI audio detection tool Aurigin. The tool’s assessment indicated that the voice heard in the clip was artificially generated, confirming that the audio did not originate from the original speech.
Conclusion
The claim that RSS chief Mohan Bhagwat called for the saffronisation of the Indian Army is false. PTI Fact Check found that the viral video was digitally manipulated, using genuine footage from an RSS centenary event but pairing it with an AI-generated audio track. The altered video was shared online to mislead viewers by falsely attributing statements Bhagwat never made.
Related Blogs
Introduction
Advanced deepfake technology blurs the line between authentic and fake. To ascertain the credibility of the content it has become important to differentiate between genuine and manipulated or curated online content highly shared on social media platforms. AI-generated fake voice clone, videos are proliferating on the Internet and social media. There is the use of sophisticated AI algorithms that help manipulate or generate synthetic multimedia content such as audio, video and images. As a result, it has become increasingly difficult to differentiate between genuine, altered, or fake multimedia content. McAfee Corp., a well-known or popular global leader in online protection, has recently launched an AI-powered deepfake audio detection technology under Project “Mockingbird” intending to safeguard consumers against the surging threat of fabricated or AI-generated audio or voice clones to dupe people for money or unauthorisly obtaining their personal information. McAfee Corp. announced its AI-powered deepfake audio detection technology, known as Project Mockingbird, at the Consumer Electronics Show, 2024.
What is voice cloning?
To create a voice clone of anyone's, audio can be deeplyfaked, too, which closely resembles a real voice but, in actuality, is a fake voice created through deepfake technology.
Emerging Threats: Cybercriminal Exploitation of Artificial Intelligence in Identity Fraud, Voice Cloning, and Hacking Acceleration
AI is used for all kinds of things from smart tech to robotics and gaming. Cybercriminals are misusing artificial intelligence for rather nefarious reasons including voice cloning to commit cyber fraud activities. Artificial intelligence can be used to manipulate the lips of an individual so it looks like they're saying something different, it could also be used for identity fraud to make it possible to impersonate someone for a remote verification for your bank and it also makes traditional hacking more convenient. Cybercriminals have been misusing advanced technologies such as artificial intelligence, which has led to an increase in the speed and volume of cyber attacks, and that's been the theme in recent times.
Technical Analysis
To combat Audio cloning fraudulent activities, McAfee Labs has developed a robust AI model that precisely detects artificially generated audio used in videos or otherwise.
- Context-Based Recognition: Contextual assessment is used by technological devices to examine audio components in the overall setting of an audio. It improves the model's capacity to recognise discrepancies suggestive of artificial intelligence-generated audio by evaluating its surroundings information.
- Conductual Examination: Psychological detection techniques examine linguistic habits and subtleties, concentrating on departures from typical individual behaviour. Examining speech patterns, tempo, and pronunciation enables the model to identify artificially or synthetically produced material.
- Classification Models: Auditory components are categorised by categorisation algorithms for detection according to established traits of human communication. The technology differentiates between real and artificial intelligence-synthesized voices by comparing them against an extensive library of legitimate human speech features.
- Accuracy Outcomes: McAfee Labs' deepfake voice recognition solution, which boasts an impressive ninety per cent success rate, is based on a combined approach incorporating psychological, context-specific, and categorised identification models. Through examining audio components in the larger video context and examining speech characteristics, such as intonation, rhythm, and pronunciation, the system can identify discrepancies that could be signs of artificial intelligence-produced audio. Categorical models make an additional contribution by classifying audio information according to characteristics of known human speech. This all-encompassing strategy is essential for precisely recognising and reducing the risks connected to AI-generated audio data, offering a strong barrier against the growing danger of deepfake situations.
- Application Instances: The technique protects against various harmful programs, such as celebrity voice-cloning fraud and misleading content about important subjects.
Conclusion
It is important to foster ethical and responsible consumption of technology. Awareness of common uses of artificial intelligence is a first step toward broader public engagement with debates about the appropriate role and boundaries for AI. Project Mockingbird by Macafee employs AI-driven deepfake audio detection to safeguard against cyber criminals who are using fabricated AI-generated audio for scams and manipulating the public image of notable figures, protecting consumers from financial and personal information risks.
References:
- https://www.cnbctv18.com/technology/mcafee-deepfake-audio-detection-technology-against-rise-in-ai-generated-misinformation-18740471.htm
- https://www.thehindubusinessline.com/info-tech/mcafee-unveils-advanced-deepfake-audio-detection-technology/article67718951.ece
- https://lifestyle.livemint.com/smart-living/innovation/ces-2024-mcafee-ai-technology-audio-project-mockingbird-111704714835601.html
- https://news.abplive.com/fact-check/audio-deepfakes-adding-to-cacophony-of-online-misinformation-abpp-1654724
On 12 August 2026, President Donald Trump signed a National Security Presidential Memorandum titled Expanding Capabilities to Combat Transnational Cyber Enabled Crime. Stripped of its bureaucratic packaging, the document does something American law has resisted for three decades: it lets private companies, under close federal supervision, break into the systems of foreign criminal networks and, in some cases, disrupt or damage them.
That is a genuinely large policy shift, even if the memorandum itself is careful, almost defensive, about how it frames the shift. Understanding why requires separating what the text actually authorizes from the "hack back" headline that has attached itself to the story within days of signing.
The problem the memo says it is solving
The White House frames this as a response to scale, not ideology. Americans reported losing more than 20.8 billion dollars to cyber enabled crime in 2025, a sharp jump from the roughly 12.5 billion dollar figure cited when the administration's earlier March 2026 executive order on cybercrime, fraud, and predatory schemes was signed. Ransomware, phishing, financial fraud, sextortion, and impersonation scams sit at the center of that number, and the administration's own supporting material points to a grim detail buried in the numbers: one in seven young people who experienced sextortion as a minor reported harming themselves as a result.
Those crimes, the memorandum argues, are increasingly the work of organized, transnational groups operating from jurisdictions the FBI simply cannot reach. Domestic law enforcement, built for domestic crime, is structurally mismatched to a threat that lives across borders and inside encrypted infrastructure. The administration's answer is to formally recruit the resource it says is best positioned to close that gap: the American cybersecurity industry itself, which the memo describes as "the most innovative and technologically advanced in the world."
What the Program actually authorizes
The memorandum directs the National Coordination Center, a body first stood up under a 2025 executive order, to build a formal Program through which vetted Participating Companies can conduct two categories of activity against foreign Cyber Enabled Transnational Criminal Organizations, defined in the text as CE TCOs.
Cyber Surveillance Operations cover unauthorized access to a target's systems for the primary purpose of collecting information or intelligence, undertaken with the intent to remain undetected. Cyber Effects Operations go further: manipulating, disrupting, denying, degrading, or destroying information systems, the infrastructure those systems control, or the data resident on them.
Crucially, CE TCOs are defined narrowly. A foreign group only counts if it targets the US government, US persons, or US interests, and it must not be an institutional arm of a foreign state or wholly directed by one. The memo builds in a presumption of innocence at the state level too: a group is assumed not to be state controlled unless clear intelligence establishes otherwise. That distinction matters enormously, because it is the line meant to separate this Program from anything resembling private warfare against a nation state.
No operation happens unilaterally. Every proposed action must be approved in writing by two Program Executive Directors, one designated by the Attorney General and one by the Secretary of Homeland Security, coordinating with each other before signing off. Participating Companies must pass what the memo calls rigorous vetting, sign contractual agreements with DOJ or DHS, and in many cases post a bond or escrow of at least one million dollars, forfeited if they breach their agreement. Within 60 days of the memo's signing, the Program Executive Directors must publish detailed operating procedures covering everything from target adjudication to what happens if an operation accidentally hits a US person's system, in which case the company must stop, run minimization procedures, and immediately notify the Center.
There is also a hard ceiling built into the design. Operations expected to cause loss of life, serious injury, or conduct that would rise to the level of a use of force or armed attack under international law, termed Critical Outcomes in the text, cannot be approved by the Program Executive Directors at all. That ceiling is the memo's clearest attempt to keep this inside the bounds of law enforcement rather than sliding into something closer to conflict.
Multiple law firms tracking the rollout, including Wiley, have been explicit on one point worth repeating because so much coverage has blurred it: this is not a green light for companies to hack back on their own initiative. Every operation remains, on paper, an act of the federal government, merely executed through a contracted private hand.
Why experts are not popping champagne
Legal caution has not stopped a wave of professional anxiety. Cyber policy veterans interviewed by outlets like CyberScoop describe the memo as a genuine philosophical break in how Washington thinks about offense in cyberspace, and the debate that followed split fairly evenly between cautious optimism and open alarm.
The core worry, echoed across nearly every serious critique, is attribution. Cyber operations are hard to trace precisely because criminals exploit shared infrastructure, proxies, and compromised third party systems to hide, and that same fog does not lift just because a government contract sits behind the operator. A former senior CISA official, Michael Garcia, put the risk plainly: pressure to attribute quickly could push companies toward lower certainty judgments about who they are actually striking, with a realistic chance of hitting the wrong server, or worse, infrastructure tied to a foreign government rather than a criminal gang. A former Cyber Command official was blunter still, describing parts of the memo on social media as a structure that could reward companies for manufacturing billable threats rather than resolving them efficiently.
Paul Rosenzweig, a former DHS policy official, raised a separate and arguably more durable problem: jurisdiction. Whatever this memo authorizes under American law, the systems being accessed usually sit inside someone else's sovereign territory, governed by that country's own criminal statutes. Washington cannot legislate away a foreign hacking law simply by calling the American company that broke it a Participating Company.
None of this makes the memo indefensible. Supporters point out, correctly, that the private sector already does enormous amounts of active defense and threat disruption work informally, through botnet takedown litigation and coordinated infrastructure seizures, and that formalizing federal oversight over that activity is arguably safer than the current improvisation. The honest position, and probably the fair one, is that the memo trades one set of risks for another, and which set turns out worse will depend entirely on the operating procedures due inside sixty days, procedures the public has not yet seen.
CyberPeace Insights: what this means beyond America's borders
A significant share of the CE-TCO activity this memo is built to target, the ransomware crews, romance investment fraud operations, and sextortion rings running out of Southeast Asia, does not victimize Americans in isolation. The scam compounds clustered along the Myanmar, Cambodia, and Laos borders, repeatedly raided over the past two years, have held thousands of trafficked workers of dozens of nationalities, Indians consistently among the largest groups rescued, alongside Chinese, Filipino, and Malaysian nationals. India has run its own repatriation efforts out of Mae Sot in Thailand, bringing citizens home several hundred at a time, and has built its own institutional response to this threat through the Indian Cyber Crime Coordination Centre, which coordinates cybercrime enforcement across states and increasingly across borders.
That shared exposure gives India and the United States real common ground here. The criminal infrastructure this American Program is designed to disrupt is, in significant part, the same infrastructure that has trafficked and defrauded Indian citizens, which gives New Delhi genuine reason to watch this experiment closely and constructively. At the same time, the Program's underlying model, private companies conducting cross-border operations under one nation's legal authorization, is a genuinely new template in international cyber governance, and how it performs over its first year will likely shape how other major digital economies, India included, think about calibrating their own frameworks for public-private cooperation against transnational cybercrime. India and other nations should closely watch whether this becomes a template worth adapting or a cautionary tale worth avoiding.
It is pertinent to note that Justice and Homeland Security departments have 60 days to write detailed operating procedures covering everything from a target-vetting rubric to a classified operational workflow and 180 days to deliver the first status report to the White House.
References
- The White House. "Expanding Capabilities to Combat Transnational Cyber Enabled Crime." 12 August 2026. https://www.whitehouse.gov/presidential-actions/2026/08/expanding-capabilities-to-combat-transnational-cyber-enabled-crime/
- Wiley Rein LLP. "Navigating the New Presidential Memorandum on Transnational Cyber Enabled Crime." August 2026. https://www.wiley.law/alert-Navigating-the-New-Presidential-Memorandum-on-Transnational-Cyber-Enabled-Crime
- SecureWorld. "Trump Memo Lets Private Firms Hack Back at Cybercriminals." August 2026. https://www.secureworld.io/industry-news/trump-authorizes-private-firms-offensive-cyber-operations
- CyberScoop. "A bold new strategy or a dangerous precedent? Experts are divided on Trump's memo." August 2026. https://cyberscoop.com/private-sector-hacking-presidential-memo-cybersecurity/
- CyberScoop. "Trump turns to private sector in offensive hacking operations memo." August 2026. https://cyberscoop.com/trump-memo-private-sector-offensive-hacking/
- CNN Politics. "Cyber privateers: Trump issues order allowing US companies to hack overseas groups under certain conditions." August 2026. https://www.cnn.com/2026/08/13/politics/cyber-privateers-trump-order-overseas-groups-hacking
- NPR. "Trump administration wants to allow companies to hack foreign cybercriminals." August 2026. https://www.npr.org/2026/08/15/nx-s1-5930311/trump-companies-hack-foreign-cybercriminals
- The Next Web. "President Donald Trump signs memo letting US agencies hack transnational crime groups abroad." August 2026. https://thenextweb.com/news/trump-cyber-memo-transnational-crime
- Machine News. "Security firms hit back at Trump's call to hack back against international crime gangs." August 2026. https://www.machine.news/security-firms-hit-back-at-trumps-call-to-hack-back-against-international-crime-gangs/
- Lawfare. "Trump Admin Cyber Strategy Centers Private Sector in Offensive Cyber Operations." March 2026. https://www.lawfaremedia.org/article/trump-admin-cyber-strategy-centers-private-sector-in-offensive-cyber-operations
- Lawfare. "Partners or Provocateurs? Private Sector Involvement in Offensive Cyber Operations." July 2025. https://www.lawfaremedia.org/article/partners-or-provocateurs--private-sector-involvement-in-offensive-cyber-operations
- Global Indian Network. "Pig Butchering Scams in India: The Dark Intersection of Social Media, AI, and Emotional Manipulation." January 2026. https://globalindiannetwork.com/pig-butchering-scams-in-india/
- The Tribune. "India brings home scammed 549 nationals from Myanmar in 2 days." 2025. https://www.tribuneindia.com/news/india/india-brings-home-scammed-549-nationals-from-myanmar-in-2-days
- Malay Mail. "India to repatriate 500 nationals fleeing Myanmar's cyber scam hub, says Thai PM." October 2025. https://www.malaymail.com/amp/news/world/2025/10/29/india-to-repatriate-500-nationals-fleeing-myanmars-cyber-scam-hub-says-thai-pm/196399
- NBC News. "260 foreigners rescued from virtual slavery in Myanmar's online scam centers are being repatriated." 2025. https://www.nbcnews.com/news/world/260-foreigners-rescued-virtual-slavery-myanmars-online-scam-centers-ar-rcna192180
- CyberPeace Foundation. "About Us." https://cyberpeace.org/about-us
Contributors
- Maj. Vineet Kumar, Founder & Global President, CyberPeace
- Mr. Neeraj Soni, Senior Research Analyst, Policy & Advocacy, CyberPeace
List of Abbreviations
- CE‑TCO — Cyber Enabled Transnational Criminal Organization
- DOJ — Department of Justice
- DHS — Department of Homeland Security
- FBI — Federal Bureau of Investigation
- CISA — Cybersecurity and Infrastructure Security Agency
- I4C — Indian Cyber Crime Coordination Centre
- US — United States
- IT — Information Technology
Executive Summary
A purported video of Hindi news channel Aaj Tak anchor Anjana Om Kashyap is being widely shared on social media. In the viral clip, Kashyap appears to be apologising and questioning her own journalistic credibility. The CyberPeace Research Wing research found the claim to be false. The probe revealed that the viral video was created by manipulating an original video of Anjana Om Kashyap. Her voice and statements were altered to falsely portray her as issuing an apology, whereas she made no such remarks in the original footage.
Claim:
A YouTube user shared the viral video claiming that Anjana Om Kashyap was apologising. The post can be seen here:
Fact Check
The keyframes of the viral video were analysed using reverse image search. During the research , the original reel was found on Anjana Om Kashyap’s Instagram account, where it was posted on December 26, 2023. In the original reel, Kashyap is seen praising Bihar. At no point does she apologise or make any statements similar to those heard in the viral clip.
Further research led to the original version of the video on Anjana Om Kashyap’s YouTube channel, uploaded on October 17, 2023. According to the video description, the speech was delivered during the ‘Bihar Meet’ event organised by the Indian People Forum in the UAE. Notably, none of the statements heard in the viral clip appear in the original speech.
The viral video was also analysed using the AI-detection tool Hive Moderation, which indicated a high likelihood of AI-generated or deepfake manipulation.
Conclusion
The research found that the viral video has been digitally altered and falsely shared on social media. The original video of Anjana Om Kashyap was edited, and the audio was manipulated to create the misleading impression that she was apologising. No such statement was made by her in the authentic video.
