Pakistan-linked social media handles target Indian institutions with warfare propaganda in 2025

Hyderabad: Since May 7, following Operation Sindoor, NewsMeter has identified a sustained and escalating pattern of AI-manipulated and AI-generated content targeting India’s political leadership, military establishment, journalists and civil society figures.

These cases are not isolated incidents but part of a coordinated misinformation ecosystem, primarily amplified via X (formerly Twitter).

In 2025, NewsMeter flagged a total of 120 instances of AI-generated content amplifying misinformation. However, between May and December 2025, there was a spike, particularly AI-manipulated videos from a recurring set of accounts, building narratives, which suggests the systematic use of Artificial Intelligence as a tool of influence and psychological warfare.

Some of the major one we found include four fully AI-generated media content—two videos depict PM Modi and one video shows Amit Shah apologising for conflicts with Pakistan; a fake AI image showing a meeting between NSA Ajit Doval and exiled Putin critic Mikhail Khodorkovsky in London that never happened; and a AI-generated song with a fictional artist, Sikh singer, BV Singh, mocking journalist.

What kind of AI content is being propagated?

Pakistani X accounts based in that country or hiding behind a South Asia location have been peddling four main types of AI-generated content:

1. Military defeat narratives: Deepfakes claiming Indian jets were shot down, airbases destroyed or that India is behind Pakistan technologically. Some even fake senior military officers admitting to failures. These tend to spike right after defence news breaks.

2. Political betrayal content: AI videos falsely showing the Prime Minister, Home Minister or Defence Minister apologising to Pakistan or reportedly agreeing to give up territory like Arunachal Pradesh or Ladakh to China. The goal is clearly to trash public confidence in leadership during security crises.

3. Internal chaos claims: Fake videos alleging arrests or persecution of activists like Sonam Wangchuk, minority oppression and election fear-mongering designed to polarise voters.

4. Diplomatic isolation narratives: Content falsely suggesting India begged for foreign help after Operation Sindoor or that Pakistan’s elections are more legitimate than ours, with the use of Congress leader Rahul Gandhi. Some fabricate statements from foreign officials to make India look weak and isolated internationally.

Recurring targets

Our analysis reveals that AI manipulators consistently target people with institutional authority and public trust.

Political leaders, including Prime Minister Narendra Modi, Home Minister Amit Shah, Defence Minister Rajnath Singh, Uttar Pradesh Chief Minister Yogi Adityanath, President Droupadi Murmu, Leader of Opposition Rahul Gandhi and Congress leader Sachin Pilot, are shown in manipulated videos making statements that they never made.

Military brass such as Army Chief Gen. Dwivedi, Navy Chief, CDS Anil Chauhan, Air Force Chief AP Singh, and retired generals like KJS Dhillon are shown blaming the government for alleged restraint during Operation Sindoor or alleged loss of Rafale jets.

Journalists like Ravish Kumar are depicted explaining the nexus between BJP leaders and Pakistan’s Imran Khan party, and Palki Sharma and Shiv Aroor are shown allegedly reporting the murder of Bangladesh student leader Osman Hadi and the terrorist attack on Bondi Beach.

Activists like Sonam Wangchuk and regional officials from Ladakh are deep-faked to push internal unrest stories, while even Afghanistan’s Foreign Minister Amir Khan Muttaqi has been falsely depicted announcing Hindu temples in Afghanistan.

Real videos are being tweaked using AI, causing major confusion

Out of 32 misinformation cases NewsMeter examined after Operation Sindoor, 26 were AI-manipulated videos, a sophisticated production in which real footage is merged with synthetic audio with voice cloning and lip-syncing.

That makes people way harder to spot than something obviously generated from scratch.

The propaganda actors sitting outside India are spreading ‘believable’ fake news by tweaking real footage rather than creating everything synthetically. An altered reality is more believable than pure fiction, and it spreads faster because people may not question what looks real.

An analysis of X accounts sharing AI content

An analysis of X accounts spreading AI-manipulated content reveals a geographically concentrated, network-driven operation.

Using VPNs: X’s new feature, allowing users to examine account locations by clicking on joining dates, has enabled us to verify these patterns. However, we have also observed that propaganda accounts are using VPNs to obscure their true locations.

Finding real location: Two accounts, The Whistle Blower (@InsiderWB) and Abubakar Qassam (@abubakarqassam), emerged as the top spreaders of AI-manipulated videos in our findings, and both frequently change their displayed locations. When clicking on their joining dates, these accounts currently appear to be based in the United States, though we previously found them listed as South Asia-based. Also, Abubakar Qassam’s account shows a connection to the Pakistan App Store, a clear indication that the person behind the account is from Pakistan.

X has cautioned that location information may not be accurate, noting that both accounts could be using proxies such as VPNs, which can alter the country or region displayed on their profiles.

Strong links to Pakistan: A significant portion of the accounts either self-identify as Pakistan-based or display clear Pakistan-linked indicators, including the Pakistani flag in cover images, distinctive posting patterns, narrative alignment and consistent amplification of the same themes.

Another group involves South Asia-based accounts (likely concealing their locations using VPNs) but consistently promotes Pakistan-centric content. A smaller subset claims locations outside South Asia.

Here is the list of X accounts flagged by NewsMeter in its stories.

NewsMeter’s findings conclude a coordinated cross-border misinformation operation running after Operation Sindoor.

AI is being weaponised to exploit the credibility of India’s trusted leaders, military officials, journalists and activists. The same persistent network of X accounts handles keeps showing up, pushing the same narratives with strategically timed postings. We have found that this is not random misinformation sharing; it is coordinated behaviour.