This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
A deepfake video of actress Jennifer Aniston was circulated on Facebook to fraudulently promote collagen supplements. The video used footage from a previous interview, with the audio replaced by an AI-generated voice clone. The incident highlights the growing prevalence of unauthorized AI-generated celebrity endorsements and the challenges of enforcing digital likeness protections.
A purportedly AI-generated deepfake video featuring Jennifer Aniston falsely promoting collagen supplements reportedly circulated on Facebook, misleading viewers about her involvement. The video, reportedly created without her consent, used footage from a previous roundtable interview, purportedly modified by AI to advertise health products.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
Using AI systems to gain a personal advantage over others such as through cheating, fraud, scams, blackmail or targeted manipulation of beliefs or behavior. Examples include AI-facilitated plagiarism for research or education, impersonating a trusted or fake individual for illegitimate financial benefit, or creating humiliating or sexual imagery.
Also classified as
The deepfake video spread false health and wellness claims, spouting questionable science about carbs, cardio, and fat burning.
The incident involved scammers using AI-generated voice clones and deepfakes of a celebrity's likeness to deceive consumers into purchasing collagen supplements for financial gain.
Human
Due to a decision or action made by humans
Intentional
Due to an expected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed
The incident was caused by human scammers who intentionally deployed an AI voice clone and deepfake video to deceive social media users for financial gain.
Added categories peak at 3: Substantial (Epistemic Harm, Inferred)
An unauthorized deepfake video of actress Jennifer Aniston was used in a commercial scam on Facebook to promote wellness supplements. The incident represents a common consumer fraud issue using accessible AI tools, but carries negligible national security implications across all assessed categories.
The AI system was used as a tool by human scammers to generate a synthetic voice clone and manipulate video footage. The AI did not exhibit autonomous planning or decision-making capabilities.
Limited Risk: The report describes the creation of an AI-generated deepfake, which falls under the category of systems requiring specific transparency obligations to ensure users are informed they are interacting with AI-generated content.
The AI system was primarily used to generate a deepfake video by lip-syncing existing footage and cloning Jennifer Aniston's voice to make it appear as though she was endorsing a product.
The reports provide clear and consistent details about the nature of the deepfake, the source footage used, the platform where it was hosted, and the approximate view count. The role of the AI as a tool for voice cloning and lip-syncing is explicitly described.
None. The incident is clearly a commercial deepfake scam.
Technology
Highly Accessible Deepfake Tools
Low-cost AI digital tools allow anyone to easily clone voices and create fakes.
Imperfect Lip Sync and Audio
The AI audio did not perfectly sync with the original roundtable video lips.
AI Voice Cloning Capability
AI cloned the celebrity's voice accurately to mimic her real speaking style.
Data Inputs
Unauthorised Use of Video
Scammers used publicly available Hollywood Reporter roundtable video footage.
Celebrity Voice Samples
Abundant public audio of the celebrity was used to train the voice clone.
Human Factors
Consumer Susceptibility
Casual social media users are easily misled by realistic celebrity deepfakes.
Exploitation of Celebrity Trust
Scammers exploited the celebrity's reputation among midlife women.
Overreliance on Public Reports
Platforms rely on users and influencers to flag inauthentic content.
Process and Methods
Reactive Content Moderation
Platforms act reactively after a video goes viral, like a game of whack-a-mole.
Delayed Takedown Response
The deepfake reached up to one million views before being flagged and removed.
Inadequate Proactive Detection
Facebook failed to automatically detect unauthorized use of celebrity likeness.
Management
Reliance on Manual Reporting
Platform management relied on user flags instead of proactive ad screening.
Lack of Proactive Ad Policies
Management failed to enforce strict verification for paid advertisements.
Regulatory Environment
Enforcement Challenges
Existing laws are difficult to enforce when creators are hard to track down.
Veto of AI Guardrail Legislation
The veto of SB 1047 left gaps in systemic oversight of AI developers.
Inconsistent Regional Laws
Varying state laws create a fragmented regulatory landscape for AI safety.
AI system involved
An AI system was involved, but the reports do not identify it.
Identified by automated extraction; can differ from the alleged developers above.
