Reliable health information is based on solid scientific evidence, up-to-date and consistent with the medical practice of reference hospital services. This definition, seemingly simple, poses a concrete problem: most health content consulted online does not meet these criteria. Medical misinformation circulates at a speed that scientific publications cannot keep up with, and its consequences range from treatment being stopped too early to consultations being delayed by several months.
Medical deepfakes and AI-generated fake doctors: the recent threat
Competitors approach health fake news from the angle of traditional social networks. A more recent phenomenon deserves particular attention: AI-generated medical deepfakes.
The European Commission estimated that there would be around 8 million deepfakes circulating worldwide in 2025, compared to 500,000 in 2023, according to a Euractiv survey published in August 2026. Among these manipulated contents are videos targeting vaccination and health agencies like the EMA.
A reader accustomed to verifying textual sources may find themselves disarmed in front of a video where a human-like face recites a structured speech. To cross-check reliable information on Skeptic North, one must now apply the same critical spirit to video content as to written articles.
The European AI Act, with some transparency obligations coming into effect from August 2, 2026, mandates labeling of content generated or modified by AI that could cause significant harm. This regulatory framework is new, and its concrete effects remain to be observed, but it marks a turning point in the regulation of misleading health content.
Cognitive biases that fuel false health beliefs
Health misinformation does not work solely because the sources are poor. It works because our brain cooperates.
Confusing popularity with reliability remains the most common trap. Content shared thousands of times on social media acquires an appearance of legitimacy. The number of shares says nothing about the scientific rigor of a claim.
Three cognitive mechanisms reinforce adherence to false medical information:
- The confirmation bias: we more easily remember information that supports what we already believe, and we dismiss information that contradicts our convictions about vaccines, diseases, or treatments.
- The repetition effect: a claim read multiple times eventually seems true, even without evidence. Social media amplify this effect by exposing the same content in different formats.
- The appeal to emotion: alarmist or overly reassuring content triggers immediate reactions. A poignant personal testimony about a disease often weighs more than a rigorous study in the reader’s mind.
Recognizing these mechanisms is not enough to protect oneself, but it allows for a pause before sharing or acting on the basis of unverified information.
Checklist for verifying online health content
Rather than a list of generic good reflexes, a technical checklist allows for a quick assessment of the reliability of medical content. Each criterion acts as a filter: if the content fails on just one point, caution is warranted.
The author must be identifiable and qualified. An article signed by a doctor, a researcher from Inserm, or an identified health professional holds more value than an anonymous text or one signed by a “wellness expert.” Doctors surveyed in a study by The Physicians Foundation cite social media, influencers, and word-of-mouth as the main drivers of misinformation encountered in consultations.
The publication date matters as much as the content. Medical knowledge evolves, and an article on vaccination from several years ago may contain outdated recommendations. Public bodies like the Haute Autorité de Santé (HAS) or Inserm regularly update their publications.
The tone of the content is a reliable indicator. A rigorous scientific article does not promise a “miracle cure” and does not seek to provoke fear. An alarmist tone or absolute promises almost always signal unreliable content.
Check if content is trying to sell something
Many sites disseminate medical information aimed at promoting a dietary supplement, a program, or an alternative method. This conflict of interest is not always visible. Looking for legal mentions, identifying the site’s funder, and spotting commercial links embedded in the content helps distinguish information from promotion.
Priority French institutional sources to consult
Evidence-based medicine produces resources accessible to the general public. Knowing them prevents dependence on social media algorithms for health information.
- The Haute Autorité de Santé (HAS) publishes good practice recommendations and treatment evaluations, available for free online.
- Inserm offers scientific popularization files on diseases, vaccines, and treatments, written by researchers.
- The site sante.fr, edited by public authorities, centralizes information validated by health professionals on common pathologies and prevention.
- Professional and patient associations publish content reviewed by specialists, often more accessible than raw scientific publications.
Becoming validated information takes time. A preliminary result reported by the media does not hold the same value as a study published in a peer-reviewed journal. This temporal gap between discovery and validation explains why health fake news always arrives before corrections.
In the face of doubt about a treatment or disease, the health professional remains the most reliable filter. Doctors observe an increase in patients influenced by misinformation during consultations, making the doctor-patient dialogue all the more necessary to deconstruct false beliefs before they translate into risky medical decisions.
