Purported Deepfake Disinformation Aired on Russian State TV Allegedly Linking Ukraine to Moscow Attack

Russian state television reportedly aired a fabricated video that appeared to imitate senior Ukrainian security officials and present a false admission of involvement in the Crocus City Hall massacre. The clip reportedly used elements from recent interviews combined with allegedly AI-generated audio to create a misleading confession. ISIS-K is reported to have publicly claimed responsibility, while independent analysts and governments have rejected the narrative promoted through the video.

Risk classification

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

The deepfake was broadcast on Russian state television as part of a coordinated disinformation campaign aimed at manipulating public opinion and influencing geopolitical dynamics.

Additional risk subdomains

• 3.1 False or misleading information: The AI-generated video and audio fabricated false statements from a public official to mislead the public about a major terrorist event.

Causal factors

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

The incident was caused by human actors who intentionally created and broadcasted a deepfake video using deployed AI tools to spread political disinformation.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Limited Risk: The report describes an AI-generated deepfake video, which falls under the transparency obligations of Risk Level 3.

AI system and alleged parties

• AI system: None named
• AI purpose: Deepfake Video Generation; Voice Generation
• Behaviour type: Tool
• Alleged developer: Russian state media, NTV Channel
• Alleged deployer: Russian state media, NTV Channel
• Alleged harmed parties: Ukraine, Truth, Oleksiy Danilov, National security and intelligence stakeholders, Media integrity, Kyrylo Budanov, Journalistic integrity, General public of Ukraine, General public of Russia, General public, Epistemic integrity

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Severe. 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 Severe, indirect Negligible
• 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 Severe, indirect Severe
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Malicious content

Reported: Yes, the reports describe the creation and broadcast of a deepfake video containing malicious, fabricated audio.

Directly caused: The deepfake video featured AI-generated audio of Oleksiy Danilov falsely claiming responsibility for a terrorist attack, broadcast on Russian state TV NTV to millions of viewers.

Inferred additional harm: The video likely spread further across social media platforms, exposing millions more to the malicious content.

Democracy

Reported: Yes, the report describes the deepfake as part of a disinformation campaign to manipulate public opinion and justify military escalation.

Directly caused: The deepfake was used by Russian state media to manipulate public perception regarding the perpetrators of the Moscow concert hall attack.

Indirectly caused: The disinformation aimed to drum up domestic support for Russia's war in Ukraine and discredit Ukraine internationally.

Inferred additional harm: Such state-sponsored deepfakes erode international trust and democratic discourse on a global scale.

Epistemic

Reported: Yes, the report describes the deepfake as adding to a 'swirl of disinformation' and sowing 'fury and confusion'.

Directly caused: The video fabricated evidence of a Ukrainian official admitting to a terrorist attack, misleading viewers who watched the NTV broadcast.

Indirectly caused: The fabrication contributed to the erosion of shared truth regarding the actual perpetrators (ISIS-K) of the attack.

Inferred additional harm: Widespread distribution of the deepfake online further pollutes the information ecosystem, making it harder for the public to distinguish real news from state-sponsored fabrications.

People affected

• Occurrences reported: 1
• People reportedly harmed: 1
• People reportedly exposed: 10000000

Potential causes

Management

• Prioritizing Narrative: Management prioritized political narratives over journalistic truth.

Technology

• AI Voice Cloning: AI cloned Danilov's voice using audio from recent interviews.
• Video Mashup Software: Software allowed mashing of two different video interviews into one.

Data Inputs

• Source Video Availability: Recent high-quality TV interviews provided the source data for the deepfake.

Human Factors

• Disinformation Intent: Malicious actors created fake content to manipulate public perception.
• Public Susceptibility: Indoctrinated audiences are highly susceptible to believable deepfakes.

Process and Methods

• No Authentication Process: Broadcasters failed to verify the authenticity of the video before airing.
• Rapid Content Assembly: Quick generation of deepfakes allows rapid response to real-world events.

Regulatory Environment

• Unregulated AI Media: No regulatory framework prevents state media from broadcasting deepfakes.

Information quality

• Classification confidence: High
• Reason for confidence: The reports from multiple reputable sources consistently describe the broadcast of the deepfake, its content, and how it was debunked by experts like BBC Verify.
• Ambiguities identified: The specific AI tools or software used to generate the deepfake are not identified.
• Alternative interpretations: None. The consensus is clear that the video was a fabricated deepfake.

Russian state television broadcasted a deepfake of Ukrainian security official Oleksiy Danilov falsely claiming responsibility for the Crocus City Hall terrorist attack. The incident represents a substantial information warfare operation aimed at manipulating public opinion and justifying further military aggression against Ukraine, highlighted by rapid international efforts to debunk the fabrication.

• Overall national security impact: Substantial
• Scope: Multiple nations
• Primary target: Ukraine
• Other affected: Russia
• Alleged perpetrator: Russian state media

Threat characteristics

• Imminence: Immediate. The incident occurred during an active crisis following a major terrorist attack, requiring rapid response from international observers and Ukrainian officials to debunk the fake.
• Autonomy: Human-controlled. The AI system was used strictly as a tool by human creators to generate synthetic audio and edit the composite video.
• Novelty: Evolved capability. While deepfakes are an established threat, broadcasting AI-generated audio on state television during an active national security crisis to justify military escalation represents an evolved threat vector.

Impact by dimension

• Physical security: Negligible. No physical attacks, autonomous weapons deployment, or critical infrastructure manipulation occurred as a direct result of this incident.
• Information security: Substantial. State-sponsored deepfake of a senior Ukrainian official was broadcast on Russian state TV to falsely attribute a major terrorist attack, representing a coordinated disinformation campaign.
• Sovereignty: Minor. The incident aimed to discredit Ukrainian government leadership and justify further Russian military aggression, though it did not directly compromise Ukraine's internal operations.
• Economic security: Negligible. No financial systems, strategic technology, or critical economic supply chains were affected by this deepfake broadcast.
• Societal stability: Minor. The deepfake was used to whip up anti-Ukrainian hysteria and manipulate public perception, risking increased hostility and social tension within the context of the ongoing war.