Incident Overview
In January 2026, fabricated videos featuring NHS clinicians from Guy’s and St Thomas’ NHS Foundation Trust, a UK public-health academic, and Oprah Winfrey were used to promote weight-loss patches and wellness supplements. These deepfake videos circulated primarily on Facebook and TikTok, impersonating trusted medical professionals to endorse unproven products. The NHS trust issued warnings about the fraudulent content, confirming its authenticity as a coordinated scam targeting vulnerable individuals seeking medical or wellness advice.
One incident involved a Utah woman who purchased weight-loss supplements after seeing a deepfake video featuring Oprah Winfrey, only to discover the product contained misrepresented ingredients, primarily turmeric, rather than the advertised formula. Similar campaigns in the UK and globally have targeted health experts, including Professor David Taylor-Robinson, to promote misleading claims about menopause and other health conditions.
The scams relied on synthetic media generation technologies, including deepfake and synthetic audio tools, to create hyper-realistic impersonations. These technologies were deployed by scammers to bypass verification processes on social media platforms, amplifying the reach of deceptive advertisements. The misuse of AI-generated content in this context highlights broader risks to epistemic integrity, as misinformation spreads rapidly through digital channels, undermining public trust in credible sources of health information.
Guy’s and St Thomas’ NHS Foundation Trust confirmed the incident, emphasizing the potential harm to patients actively searching for medical guidance. The trust’s warning underscored the growing threat posed by AI-driven fraud, particularly in healthcare, where misinformation can lead to dangerous decisions regarding personal health and financial exploitation.
This incident is part of a documented global pattern, with similar cases reported in Sweden, New Zealand, and Australia, where deepfakes of doctors and public figures were used to sell health supplements and promote misleading medical advice. The AI Incidents Database (AIID) and the OECD AI Incidents and Hazards Monitor (AIM) have classified these incidents under the domains of misinformation, malicious actors and misuse, and socioeconomic harms, reflecting the multifaceted risks posed by synthetic media in digital advertising.
The deepfake videos were designed to mimic the appearance and voice of the impersonated individuals, often altering real footage or cloning voices to spread deceptive claims. For example, the deepfake of Oprah Winfrey in the Utah case was used to lend credibility to a product advertised as scientifically backed, despite containing no such evidence. Similarly, the NHS clinicians’ impersonations included fabricated endorsements for weight-loss patches, exploiting the authority associated with medical professionals to manipulate consumer behavior.
Social media platforms played a critical role in the dissemination of these scams, as the videos were shared widely on Facebook and TikTok. Users searching for health advice or weight-loss solutions encountered these deepfakes alongside legitimate content, making it difficult to distinguish between authentic and fabricated endorsements.
The misuse of AI technologies in this incident extends beyond financial fraud to include the erosion of trust in healthcare institutions and public figures. Patients and consumers who rely on online platforms for medical information may unknowingly encounter manipulated content, leading to misinformed decisions about their health.
The Domain Taxonomy of AI Risks categorizes this incident under multiple risk domains, including misinformation, malicious actors and misuse, and socioeconomic harms. The primary cause of the risk is attributed to the development and deployment of deepfake and synthetic audio technologies by scammers, who exploited these tools to create convincing impersonations. The risk occurred as an unexpected outcome of pursuing financial gain through deceptive marketing, rather than as an intended consequence of legitimate AI development.
Possible Root Cause
The root cause of this incident lies in the convergence of accessible AI tools, malicious intent, and the vulnerabilities of social media platforms. Scammers leveraged these tools to create fabricated videos that mimicked the likeness and voice of trusted figures, including NHS clinicians, a UK public-health academic, and Oprah Winfrey.
The development of these AI systems was not inherently malicious, but their misuse by scammers enabled the creation of deceptive content at scale.
Social media platforms, particularly Facebook and TikTok, played a pivotal role in amplifying the reach of these deepfake videos. This creates an environment where deceptive advertisements can spread rapidly, often before platform moderators or fact-checkers can intervene.
The lack of robust verification mechanisms on these platforms further exacerbated the problem. By the time fraudulent videos are identified and taken down, they may have already reached a wide audience, causing harm to individuals who trust the fabricated endorsements.
Another contributing factor is the anonymity afforded to scammers operating online.
The incident also highlights the broader challenge of distinguishing between authentic and synthetic media in an era of proliferating AI-generated content. As deepfake technologies become more sophisticated, the line between real and fabricated videos continues to blur, making it increasingly difficult for consumers to verify the credibility of online information. This erosion of trust in digital media poses significant risks to public health, particularly in contexts where misinformation can lead to harmful decisions.
The misuse of AI in this scam campaign was not an isolated event but part of a growing trend documented by organizations such as the AI Incidents Database (AIID) and the OECD AI Incidents and Hazards Monitor (AIM). These entities have identified similar incidents worldwide, where deepfakes of healthcare professionals and public figures have been used to promote fraudulent products and spread misleading health advice. The pattern suggests a systematic exploitation of AI technologies by scammers to target vulnerable populations seeking medical or wellness solutions.
Takeaways
This incident underscores the urgent need for improved detection and mitigation strategies to combat the misuse of AI-generated media in fraudulent advertising. Social media platforms must enhance their content moderation capabilities, particularly in identifying and removing deepfake videos that impersonate public figures or healthcare professionals. Proactive measures, such as AI-driven detection tools and human oversight, can help reduce the spread of misleading content before it causes harm.
Regulatory frameworks must also evolve to address the ethical and legal implications of synthetic media.
Public awareness campaigns are critical in educating consumers about the risks of deepfake scams. Healthcare institutions, such as Guy’s and St Thomas’ NHS Foundation Trust, can play a key role in disseminating warnings about fabricated endorsements and providing resources to help individuals verify the authenticity of online content.
As AI-generated content becomes more prevalent, maintaining trust in credible sources of information requires transparency about the origins of media, whether real or synthetic.
For healthcare professionals and public figures, the rise of deepfake scams necessitates proactive measures to protect their identities and reputations. This includes monitoring online platforms for unauthorized use of their likeness and collaborating with law enforcement to track and dismantle fraudulent campaigns. Organizations like the NHS can issue public statements to clarify their official positions on medical treatments and supplements, countering the spread of misleading claims.
The economic and health risks associated with these scams extend beyond individual victims to broader societal impacts. Misleading advertisements for weight-loss products and wellness supplements can lead to financial losses for consumers, as well as potential health consequences from relying on unproven treatments.
The misuse of deepfake and synthetic audio tools in fraudulent advertising reflects a broader challenge in balancing innovation with ethical responsibility.
Finally, the global nature of this incident emphasizes the need for international cooperation in combating AI-related fraud.
Sources
- incidentdatabase.aihttps://incidentdatabase.ai/cite/1405/
- incidentdatabase.aihttps://incidentdatabase.ai/cite/1397/
- incidentdatabase.aihttps://incidentdatabase.ai/cite/1408/
Incident summary from the AI Incident Database, used under CC BY-SA 4.0.