Reported AI-Generated Audio of Ukrainian Commander Andriy Biletsky Used in Russian Disinformation Campaign

In late June 2025, Russian Telegram channels reportedly circulated a video containing a purportedly AI-generated audio track impersonating Ukrainian commander Andrii Biletskyi. The audio clip reportedly claimed Ukrainian authorities deliberately avoid identifying fallen soldiers to withhold compensation. Verification reportedly showed the voice was synthetic and mismatched with original May 16 footage of Biletskyi. Hive Moderation reportedly confirmed the audio was overwhelmingly AI-generated.

Risk classification

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

The incident involves a coordinated disinformation campaign on Russian Telegram channels using deepfake technology to manipulate public opinion and undermine trust in the Ukrainian government during an active conflict.

Additional risk subdomains

• 3.1 False or misleading information: The AI system was used to generate highly realistic but entirely fabricated audio of a military commander making false claims.

Causal factors

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

The risk was caused by human actors intentionally deploying and using AI voice generation technology to create and spread a political disinformation video.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Risk Level 3. Limited Risk: The report describes AI-generated content (a deepfake video), which falls under transparency obligations for synthetic media under the EU AI Act.

AI system and alleged parties

• AI system: Hive Moderation (Hive)
• AI purpose: Voice Generation; Content Verification
• Behaviour type: Tool
• Alleged developer: Unknown voice cloning technology developers, Unknown deepfake technology developers
• Alleged deployer: Russian Telegram channels, Russian disinformation channels, Government of Russia-aligned actors
• Alleged harmed parties: Truth, National security and intelligence stakeholders, Military of Ukraine, Government of Ukraine, General public of Ukraine, Families of military personnel in Ukraine, Epistemic integrity, Andriy Biletsky

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 Negligible
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Negligible, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Negligible, indirect Negligible
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Minor, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Malicious content

Reported: Yes

Directly caused: The report describes a deepfake video containing AI-generated audio mimicking Commander Andrii Biletskyi circulating on Russian Telegram channels.

Inferred additional harm: The video likely reached tens of thousands of viewers on Telegram, spreading malicious propaganda designed to undermine morale.

Epistemic

Reported: Yes

Directly caused: The report explicitly states that Russian Telegram channels circulated a fake video with AI-generated audio falsely claiming Ukrainian authorities avoid identifying fallen soldiers.

Inferred additional harm: This campaign contributes to the broader pollution of the information ecosystem, making it harder for citizens to distinguish real wartime communications from synthetic propaganda.

People affected

• Occurrences reported: 1
• People reportedly harmed: 1
• People reportedly exposed: 10000

Potential causes

Management

• Inadequate Content Moderation: Platform management failed to establish proactive deepfake filtering.
• Lack of Incident Response: No rapid response protocols to mitigate viral state-sponsored disinformation.

Technology

• AI Voice Cloning Technology: Generative AI allowed highly realistic cloning of the commander's voice.
• AI Video Editing Tools: AI tools allowed seamless integration of synthetic audio with real footage.

Data Inputs

• Publicly Available Footage: The original video from May 16 was accessible online for manipulation.
• Voice Training Data: Public audio of the commander enabled training of the voice clone.

Human Factors

• Target Audience Susceptibility: Viewer emotions are easily exploited by sensitive topics like fallen soldiers.
• Malicious Intent of Actors: State-sponsored actors actively sought to spread destabilizing propaganda.

Process and Methods

• Unmoderated Distribution Channels: Telegram channels allowed rapid, unchecked spread of the deepfake video.
• Reactive Fact-Checking: Debunking occurred only after the video had already widely circulated.

Regulatory Environment

• Lack of Deepfake Regulations: No international legal frameworks exist to deter state-backed deepfakes.
• Weak Platform Accountability: No regulatory pressure on messaging apps to instantly remove fake media.

Information quality

• Classification confidence: High
• Reason for confidence: The report clearly details the nature of the deepfake, the specific claims made, the target of the impersonation, and the verification method (Hive Moderation) used to debunk it.
• Ambiguities identified: The specific AI voice generation tool used to create the deepfake is not named.

A disinformation campaign utilized AI-generated audio to falsely depict Ukrainian commander Andrii Biletskyi claiming the government was neglecting fallen soldiers. While quickly debunked by AI detection tools, the incident highlights the ongoing exploitation of AI voice-cloning technology by foreign adversaries to conduct information warfare and undermine public trust during active conflicts.

• Overall national security impact: Substantial
• Scope: Single nation
• Primary target: Ukraine
• Alleged perpetrator: Russian actors

Threat characteristics

• Imminence: Long-term. The specific incident has been debunked, but the underlying threat of AI-enabled psychological operations remains an ongoing strategic concern.
• Autonomy: Human-controlled. The AI tool was directly controlled by human operators to generate the fake audio and integrate it into the video.
• Novelty: Established threat. AI-generated deepfakes and voice clones have been repeatedly used as propaganda tools in geopolitical conflicts, representing an established threat.

Impact by dimension

• Physical security: Negligible. The incident involved digital disinformation and did not result in physical damage, kinetic attacks, or critical infrastructure compromise.
• Information security: Substantial. The incident represents a coordinated information warfare campaign using sophisticated AI voice cloning to impersonate a military commander during an active conflict.
• Sovereignty: Minor. The campaign targeted public trust in Ukrainian military leadership and government compensation policies, though it did not disrupt core operations.
• Economic security: Negligible. There were no reported impacts on financial systems, critical supply chains, or strategic technological advantages.
• Societal stability: Minor. The disinformation targeted families of fallen soldiers to sow social discord and erode trust, though the societal impact was mitigated by rapid debunking.