Claim: A viral video shows Pakistan’s defence chief, Field Marshal Syed Asim Munir, making shocking remarks about the Operation Bunyan-un-Marsoos during the country’s May 2025 war with India, saying, “We made fools out of Pakistani citizens.”

Fact: An AI-generated deepfake audio was likely overlaid onto footage from Field Marshal Asim Munir’s address at the National Ulema and Mashaikh Conference in Islamabad in December 2025. 

 

On 23 December 2025, X (formerly Twitter) user @RealWahidaAFG posted a video showing Field Marshal Munir revealing that Pakistan’s leadership, as well as the top military brass, “made fools out of Pakistani citizens” during Operation Bunyan-un-Marsoos, which was launched to counter India’s attacks in May 2025.

His full remarks, which are in Urdu, read as follows:

“God is witness that during [Operation] Bunyan-un-Marsoos, we made fools out of Pakistani citizens. We claimed to have downed Indian jets but, I swear to God, the entire army, including President [Asif Ali Zardari], were terrified. But God forbid the people of Pakistan think for themselves.” 

@RealWahidaAFG (archive) is an apparent pro-Afghan, anti-Pakistan, and proIndia user who claims to be a “HR [human rights] defender”. We have previously debunked their false claims and other outlets accredited by the International Fact-Checking Network (IFCN) have also disproved their posts.

 

Fact or Fiction?

Soch Fact Check did not find any original footage where Asim Munir is mocking the citizens of Pakistan or saying that the army was scared during the Operation of May 2025. 

In fact, the coverage seen by verified Pakistani news channels here and here in December show that his speech was patriotic. He referred to Operation Bunyan-un-Marsoos in the context of Pakistan’s security and defence, and stated that Pakistan experienced divine help during the May 2025 conflict with India,  invoking Quranic verses to emphasise faith and unity. He stressed that jihad can only be declared by the state, rejected the legitimacy of militant violence, and urged religious scholars to counter extremism through knowledge and cohesion. Munir also accused India of sponsoring terrorism, and asserted that Pakistan confronts enemies openly rather than covertly. 

Soch Fact Check reverse-searched the images and keyframes from the viral videos to ascertain their origin and whether they were manipulated. The only exact image of this frame, was found below in this search: 

However, this reverse image search lead to the original footage: 

Deepfake detectors’ results

To further investigate if the video was altered, we ran it through DeepFake-o-Meter, an AI-based tool that detects manipulated or synthetic media, particularly deepfakes. It uses multiple detection models to analyse visual and audio cues that may indicate tampering. The results were as follows:

The second tool that was utilised for this clip is, Deepware Scanner, and the results shown below detect suspicious activity. There is an 81 percent chance of deepfake activity. 

Furthermore, Soch Fact breaks downs what these tools detect within this clip, with their definitions for deeper understanding of these results: 

The Avartify model looks for visual and temporal clues that point to AI face-swapping, reenactment, or synthesis.

The Seferbekov model uses machine learning to examine frame-level anomalies, especially in facial texture, lighting, and blending artifacts.

The Ensemble model combines the outputs of multiple models to generate a more balanced and robust prediction. The idea is to reduce false positives and increase overall reliability.

Sound Engineer’s Analysis 

Soch Fact Check reached out to Shaur Azher, a lecturer who teaches sound design and sound recording at the University of Karachi and the Shaheed Zulfikar Ali Bhutto Institute of Science and Technology (SZABIST). He also works as an audio engineer at our sister organisation, Soch Videos, and specialises in mixing and mastering audio.

Azher carried out a forensic analysis by comparing 2 samples: Sample A is the audio extracted from the video in the claim, Sample B is the audio from the original conference footage sourced from Dawn News  Army Chief General Asim Munir’s Speech From Convention for Overseas Pakistanis

Both Samples A and B appear to be recorded indoors as they contain sounds of hall reverb, Azher noted. His observations on the two samples are presented as follows:

 

  1. Frequency Spectrum Analysis

 Sample A:

  •  Frequency range: 20 Hz – 9,697 Hz
  •  Vocal energy concentrated around 2.5 kHz
  •  Spectral cluttering observed from 2.5 kHz to 9,697 Hz
  •  No natural vocal harmonics present

 Sample B:

  • Frequency range: 20 Hz – 16,000 Hz
  • Consistent and full-spectrum vocal harmonic energy
  •  Natural spectral distribution across the full audible range

 

  1. Plosive Analysis

 Sample A:

  • Plosive detected in the 20 Hz – 100 Hz range on the word “Fauj”
  • Theoretically improbable under conventional indoor recording conditions due to

microphone wind filters

 Sample B:

  •  No plosives detected

 

  1. Reverb and Acoustic Characteristics
  •  Hall reverb present in both samples, consistent with indoor hall recordings

 

  1. Vocal Tone and Dynamics

 Sample A:

  •  Monotone and compressed vocal delivery
  •  Lack of natural tonal variation

 Sample B:

  •  Natural tone with dynamic variation
  •  No compression artifacts observed

 

  1. Conclusion

The analysis indicates that Sample A exhibits characteristics inconsistent with a natural indoor recording:

  1. Limited frequency spectrum and lack of harmonic content
  2. Presence of plosive artifacts that are theoretically improbable, for example 
  3. Compressed and monotone vocal delivery.                                                                                

The dynamic range of the spoken dialogues is levelled in terms of volume and gain, As in Sample B the dynamic range is uneven, therefore it is uncompressed. 

 

Conversely, Sample B exhibits characteristics of a natural, original speech recording:

  1. Full-spectrum harmonic content
  2. Absence of unnatural plosives
  3. Natural tonal and dynamic properties                                                                                     

As in Sample B the dynamic range is uneven, therefore it is uncompressed. 

Based on these observations, Sample A was likely synthetically altered or manipulated whereas Sample B represents the authentic original recording.

The analysis indicates that Sample A exhibits characteristics inconsistent with a natural indoor recording, whereas Sample B was broadcasted from a trusted source like Dawn News

 

Virality

Soch Fact Check found the videos posted on X here and here

 

Conclusion:  AI-generated deepfake audio was likely overlaid onto footage of Field Marshal Asim Munir’s address at the National Ulema and Mashaikh Conference in Islamabad in December 2025. He does not reflect on the fear within the army or the president at any point in the original footage of the speech nor does he mock the Pakistani people for being fooled by the army. 

Subscribe
Notify of
guest
0 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
0
Would love your thoughts, please comment.x
()
x