July 31, 2026 (Conversation) Recent reports have highlighted AI-generated “doctors” spreading dubious health advice to millions of social media users. Fabricated wellness trends can travel just as quickly. A supposed “pink jelly” weight-loss treatment, for example, reportedly fooled at least one celebrity.
These examples gain credibility in different ways. Influencers can appear familiar and sincere, while AI-generated doctors can look like qualified professionals. Neither familiarity nor the appearance of expertise tells us whether a medical claim is accurate.
Turning to friends and family for health advice is understandable. They may know our circumstances in ways that a healthcare professional cannot during a brief appointment. They can provide emotional support and practical help, and their advice is usually well meant. But somebody can care about us, speak honestly and still be wrong.
Truth and trust
My own research and wider research on digital health misinformation shows that trust can play an important part in deciding which claims people accept. Trust in clinicians and scientific institutions is generally associated with lower susceptibility to false health claims. Misinformation can nevertheless spread through trusted social networks when people share it because they want to help or warn others.
Social media influencers can reproduce some of the signals of friendship without having a personal relationship with their followers. They address followers directly, share details about their lives and communicate in an informal style. Over time, followers may develop what researchers call a parasocial relationship: a one-sided sense of familiarity with a media figure. A recent meta-analysis found that parasocial relationships can influence health attitudes and behaviour, although their effects vary across different contexts.
AI-generated doctors use a different route to credibility. Accounts designed to resemble real clinicians can borrow the appearance and language of medical authority without providing any reliable way to check who created the content, what qualifications lie behind it or whether the advice is supported by evidence.
Personal stories are especially persuasive. Someone describes being ill, discovering a treatment and recovering. The account is easy to follow and often more memorable than a discussion of probabilities.
Even a completely truthful anecdote cannot show that a treatment caused the recovery or that it will be safe and effective for other people. Symptoms may have improved naturally. The person may have used several treatments at once, received an uncertain diagnosis or left out relevant details. Experiments have found that anecdotes can influence medical decisions even when people are also given stronger statistical evidence.
Fear, hope and accuracy
Emotion can add to the effect. Fear or hope may draw attention to a claim and make it feel urgent. Emotional tone does not show whether the information is accurate. Research suggests, however, that heightened emotional responses can increase susceptibility to digital health misinformation.
Identity can also shape judgment. Advice from another parent or fellow patient may appear more credible because the speaker seems to understand our experience. Rejecting the advice may then feel like rejecting a community to which we belong. A systematic review found that existing beliefs, knowledge, trust and reasoning styles can all affect how people evaluate health claims.
People may also have sound reasons for distrusting healthcare institutions. Dismissive treatment, discrimination, rushed appointments and poor explanations can damage trust. Research on trust in health organisations suggests that mistrust may be a rational response to institutional failure or unequal treatment. Telling patients to trust experts will not repair relationships that healthcare systems have damaged.
So how can we assess a health claim without dismissing personal experience?
1. Examine the claim separately from the person making it
A source may be sincere, charismatic or visually convincing without having the expertise to judge whether a treatment works. Check their relevant qualifications and whether their identity can be independently verified.
2. Look for supporting evidence and independent agreement
Do the NHS, an appropriate professional body or a reputable medical charity offer similar guidance? A single study rarely settles a medical question, so substantial claims should be supported by a wider body of research.
3. Check for commercial interests
Look for products, affiliate links, paid partnerships or private treatment programmes. Financial involvement creates a potential conflict of interest that readers should know about.
4. Be cautious about sweeping promises
Claims that a treatment produces rapid results, works for almost everyone or has no meaningful risks are unlikely to reflect the uncertainty and individual variation involved in medicine.
5. Know the warning signs
Treat secrecy claims and advice to abandon prescribed care as serious warning signs The greater the potential harm, the more important it is to check the advice with an appropriately qualified professional, such as a pharmacist or GP.
Individuals cannot carry the whole burden of detecting misinformation. Social media platforms influence which claims are amplified, while health services influence whether people feel heard and respected. Researchers are increasingly calling for a systems approach to health misinformation that combines clearer public information, responsible platform design, effective regulation and trustworthy clinical communication.
Friends and relatives can provide valuable support and personal testimony can reveal experiences that medical research may miss. Judging whether a treatment is safe and effective, however, requires evidence that extends beyond any one person’s story.