Manipulated Video Using AI-Generated Audio Targets Pakistan's Prime Minister Shehbaz Sharif with Fabricated Interview

A viral video falsely claimed to show Layal al-Ekhtiar of Al Arabiya confronting Pakistan's Prime Minister Shehbaz Sharif over funding a 30-person Umrah delegation with public money while seeking aid from Saudi Arabia. Investigators concluded the video had been dubbed with AI-generated audio. The original interview was broadcast by Saudi-owned Al Arabiya in January 2023, and it contained no such exchange. Lip-sync inconsistencies reportedly confirmed the manipulation.

Risk classification

• Primary risk domain: 4 Malicious actors
• Primary risk subdomain: 4.1 Disinformation, surveillance, and influence at scale

The deepfake was created and spread at scale to manipulate public opinion and political discourse regarding the Prime Minister's government spending.

Additional risk subdomains

• 3.1 False or misleading information: The deepfake generated false audio of a real journalist and the Prime Minister, creating highly misleading political content.

Causal factors

• Entity: Human
• Intent: Intentional
• Timing: Post-deployment

The incident was caused by a human actor intentionally using a deployed AI voice generation tool to create and spread a political deepfake.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Limited Risk: The report describes an AI-generated deepfake, which falls under Limited Risk due to transparency obligations requiring users to be informed they are interacting with or viewing AI-generated media.

AI system and alleged parties

• AI system: Attestiv.ai, ChatGpt's Deepfake Detective, Deepware, unspecified (Attestiv, Deepware, OpenAI)
• AI purpose: Voice Generation; Deepfake Video Generation
• Behaviour type: Tool
• Alleged developer: Unknown voice cloning technology developers, Unknown deepfake technology developers
• Alleged deployer: social media users, Adil Raja
• Alleged harmed parties: Shehbaz Sharif, Layal al-Ekhtiar, General public of Pakistan

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Minor. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Negligible
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Negligible, indirect Negligible
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Minor, indirect Minor
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Negligible, indirect Negligible
• Democracy: direct Minor, indirect Minor
• Privacy: direct Negligible, indirect Negligible
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Minor, indirect Minor
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Malicious content

Reported: Yes, the report describes the creation and spread of a malicious political deepfake video.

Directly caused: The deepfake video was created using AI-generated audio to falsely depict a confrontation, gaining over 731,000 views across TikTok and X.

Indirectly caused: The spread of the video fueled public criticism and political hostility against the Pakistani government during a sensitive economic period.

Inferred additional harm: It is likely that the video reached over 1,000,000 people across various social media platforms, amplifying political polarization and toxic discourse.

Democracy

Reported: Yes, the report describes the use of synthetic media to spread political disinformation targeting the Prime Minister.

Directly caused: The deepfake manipulated public perception of the Prime Minister's official travels and government spending, undermining political trust.

Indirectly caused: The video was shared by political dissidents and opposition supporters to delegitimize the government during a financial crisis.

Inferred additional harm: The incident contributes to the erosion of democratic discourse and public trust in official communications by normalizing the use of political deepfakes.

Epistemic

Reported: Yes, the report describes a fact-check that debunked the false claim and the AI-generated audio.

Directly caused: The video fabricated a non-existent confrontation between a journalist and the Prime Minister, misleading hundreds of thousands of viewers.

Indirectly caused: The viral spread of the video forced fact-checking organizations (iVerify Pakistan) to dedicate resources to verify and debunk the synthetic media.

Inferred additional harm: The widespread dissemination of highly realistic deepfakes degrades the overall information ecosystem, making it harder for citizens to distinguish real news from fabricated content.

People affected

• Occurrences reported: 1
• People reportedly harmed: 2
• People reportedly exposed: 731000

Potential causes

Management

• Engagement Prioritization: Platforms prioritized viral reach over immediate content verification.

Technology

• AI Audio Dubbing Tools: AI tools easily generate realistic fake audio dubbed over real video.
• Inconsistent AI Detection: Detection tools gave highly conflicting results from 0 to 95 percent.

Data Inputs

• Misused Real Video Footage: Authentic 2023 interview footage was manipulated with fake audio.
• Lack of Video Provenance: No digital watermarks to verify the integrity of the original media.

Human Factors

• Malicious Dissemination: Vlogger shared the fake video to manipulate public opinion.
• Confirmation Bias: Public anger over spending made users share without verifying.
• Political Motivation: Supporters shared the clip to target political opponents.

Process and Methods

• Delayed Fact Checking: Fact-check was only initiated after the fake video went viral.
• Inadequate Platform Filtering: Social media platforms failed to flag the synthetic media automatically.

Regulatory Environment

• Weak Deepfake Regulations: No clear legal frameworks to deter creation of political deepfakes.
• Lack of Platform Accountability: No regulatory enforcement to make platforms verify political content.

Information quality

• Classification confidence: High
• Reason for confidence: The reports provide clear, verified facts from a reputable fact-checking organization (iVerify Pakistan) detailing the origin of the video, the specific platforms and view counts, and the verification process showing the audio was AI-generated.
• Ambiguities identified: The specific AI tool used to generate the synthetic audio is not identified.
• Alternative interpretations: None. The evidence clearly points to an intentional political deepfake.

A political deepfake video falsely depicting Pakistani PM Shehbaz Sharif was shared widely on social media to manipulate public opinion regarding government spending. While the incident represents a minor information security threat easily managed by standard fact-checking, it highlights the evolving challenge of detecting AI-generated voice cloning and its ongoing use in political disinformation.

• Overall national security impact: Minor
• Scope: Single nation
• Primary target: Pakistan
• Alleged perpetrator: Unknown

Threat characteristics

• Imminence: Long-term. The immediate threat of this specific video has passed, but it represents an ongoing strategic concern regarding the weaponization of synthetic media.
• Autonomy: Human-controlled. The AI voice-cloning technology was used directly as a tool by human creators to generate the synthetic audio, with no autonomous system decision-making.
• Novelty: Evolved capability. While political deepfakes are established, the failure of multiple commercial AI-detection tools to consistently identify the manipulation represents an evolved challenge.

Impact by dimension

• Physical security: Negligible. No physical systems, critical infrastructure, or human safety were threatened or impacted by this incident.
• Information security: Minor. Represented a minor AI-enhanced disinformation campaign targeting national leadership, which was quickly mitigated by standard fact-checking procedures.
• Sovereignty: Minor. The deepfake targeted the Prime Minister to manipulate public perception of government spending, but did not disrupt core government operations or constitutional processes.
• Economic security: Negligible. No threats to strategic industries, financial systems, or technological competitive advantages were reported.
• Societal stability: Minor. The video aimed to fuel political hostility during an economic crisis, but did not result in mass civil unrest, systemic oppression, or human rights violations.