How food misinformation spreads:
- Vague claims and questionable certifications can mislead without being false.
- Social media and AI can amplify emotive, poorly evidenced food content.
- Producers need stronger checks, clearer sourcing and transparent sponsorship.
Consumers face food and health claims everywhere: a ‘natural’ label in a supermarket, a viral warning that an everyday ingredient is ‘secretly toxic’, an influencer promoting a ‘miracle food’ or packaging carrying a health benefit that isn’t backed by credible evidence. These messages can shape what people consider healthy, safe or trustworthy, often before they’ve checked the facts.
The EU-funded INFOODMATION project is examining this increasingly crowded landscape. Its researchers analysed 322 potentially misleading examples of food communication collected across 13 European countries, alongside more than 540,000 social media posts from popular influencers across 16 European countries. The work looks at how misinformation is generated and spread, how it affects food choices and what different stakeholders can do about it.
Pablo Hernández Escayola, head of academic research at Spanish fact-checking organisation Maldita.es in Madrid, says disinformation describes false content shared by someone who knows it’s untrue and wants to manipulate the audience. Misinformation is inaccurate material passed on by someone who believes it. Once a claim goes viral, however, proving the intention of every person sharing it is virtually impossible.
“We don’t like to point the finger at the sender of the message,” says Hernández Escayola. “We try to help the people receiving it to know whether it’s true, false or whether they need more context to interpret it correctly.” INFOODMATION therefore also examines ‘potentially misleading content’ – a much broader category that producers need to understand, even when a statement doesn’t clearly break the law.
Food claims rarely need to be entirely false
The most common problem isn’t the blatant fabrication many people associate with misinformation. Instead, questionable food communication frequently occupies an awkward space between accurate and inaccurate, according to Rocío Benavente, deputy director and sustainability and development strategy coordinator at Fundación Maldita.es in Madrid.
“It’s more about what it implies, what it suggests and what it makes you think,” she says. “It can be partially true and partially false but imply something that isn’t completely true or evidence based.”
INFOODMATION identifies three communication tactics of which producers should be particularly wary. The first involves questionable awards, endorsements or certifications that seem to carry the backing of a respected, independent authority. These stamps and badges appear on packaging and across wider brand communications, creating an impression of external scrutiny that might not withstand closer examination.
The second is advertising that resembles impartial editorial content. When a company has paid for an article, video or broadcast but the commercial relationship isn’t obvious, consumers can’t properly assess the potential conflict of interest behind what appears to be neutral information.
Highly emotional language provides the third route. Some messages alarm consumers with warnings about toxins, additives or supposedly dangerous foods. Others appeal to nostalgia, cultural identity, tradition or ideas of purity. Both approaches can guide shoppers towards a conclusion without supplying enough meaningful evidence for them to assess it.
The narratives generally fall into two broad camps. One encourages people to consume something because it’s ‘natural’, ‘local’, a ‘superfood’ or capable of ‘detoxing’ the body, and this content is often linked to selling a product. The other urges people to avoid something because it’s supposedly dangerous, toxic or packed with harmful additives – the type of alarmist message commonly used to capture attention and build an online following.
The problem, however, is that some of these ideas begin with a legitimate point. Local food may have a lower environmental impact in certain circumstances, but ‘local is always better’ can quickly turn into an exaggerated nutritional, environmental or cultural claim the evidence doesn’t support.
Crucially, legal compliance isn’t synonymous with responsible communication. INFOODMATION’s researchers haven’t identified widespread, flagrant breaches by the food industry, but they have found companies operating within ambiguous territory.
“They play very well with these lines that aren’t very clear,” says Benavente. “We have to make a call for the responsibility of the food industry to think about this and reflect on it.”
The risk stretches across packaging, websites, retailer materials, media partnerships, sponsored content and influencer briefs. A claim might survive legal review yet still damage a company’s credibility if shoppers later believe they were encouraged to infer more than the available evidence could support.
Social media and AI raise the stakes
INFOODMATION identifies social media as the main channel through which potentially misleading food content spreads. Short-form video is particularly compatible with direct, emotive and alarmist messages, although its dominance across social media more broadly also contributes to its prominence.
“Platforms work in a way that pushes and rewards emotional content that catches attention,” says Benavente. “Something very alarmist about food is going to attract attention. We eat every day, and we think this is very important for us and for taking care of our families.”
Traditional media applies professional judgement when selecting and checking stories, even if mistakes still occur. Social platforms have fewer effective gatekeepers, while algorithms tend to reward content that generates reactions rather than information that’s properly evidenced. Detailed scientific explanations must therefore compete with confident videos promising that one ingredient will cure, poison, cleanse or transform the viewer.
The effects can be more serious among consumers who are already vulnerable. INFOODMATION identifies children and adolescents, people with chronic health conditions and those of lower socioeconomic status as groups that may be particularly susceptible. Young users spend considerable time on social media, consume each piece of content quickly and may prioritise emotion, appearance or peer approval over evidence. People living with unresolved health conditions can be more receptive to foods, supplements or diets promising dramatic results. Lower-income consumers may face additional barriers involving access to healthier foods, education and food literacy.
“People who are desperate to believe something could change the way they live and solve their health problems are very vulnerable to messages promising that a food or diet will solve everything,” says Hernández Escayola.
Influencers need to appreciate the authority they can acquire with these audiences, he adds. “Young people follow lots of influencers, trust these influencers and follow their biases. It’s really important that influencers are aware of the power they have over their audience and are committed to being responsible.”
Generative AI could amplify the problem. Hernández Escayola warns that chatbots may reinforce mistaken beliefs because they’re designed to be helpful, polite and responsive to the user. When reliable evidence is limited, a model may construct a confident answer instead of acknowledging that it doesn’t know.
“They try to tell you that you’re right and that you have excellent ideas,” he says. “Sometimes, if you start with a bad idea and the chatbot gives you reasons why it’s a good idea, it can lead you to bad conclusions.”
He also cites studies that tested the same AI models using fictional users from different socioeconomic backgrounds. The accuracy of the answers varied substantially between the user profiles. Although more research is needed, the findings raise the possibility that personalisation could deliver poorer information to those already most vulnerable to misleading health claims.
Commercial influence adds another risk. Hernández Escayola highlights a recent preprint – a research paper made publicly available before it’s undergone formal peer review – that found that large language models could prioritise company incentives over user welfare. Examples included recommending a sponsored product costing almost twice as much and obscuring prices in unfavourable comparisons. The findings need to be treated cautiously until the study has been scrutinised by other experts, but they raise important questions about what could happen if paid placements are built into chatbots without clear disclosure.
It’s therefore essential that food companies treat AI-generated copy, nutrition explanations and influencer materials as drafts requiring human review. Factual assertions need traceable sources, while health, nutrition and sustainability claims must pass expert and legal checks. Hernández Escayola also advises users to request the sources behind a chatbot’s statements and verify what it provides. When checking every line isn’t practical, anything unexpected, unusually definitive or commercially convenient deserves closer scrutiny.
How producers can communicate without losing attention
Evidence-based food communication has a built-in disadvantage online because science comes with caveats, context and uncertainty, while misinformation can be compressed into a simple emotional message.
“The format is good for disinformation, misinformation and alarmism, and not so good for accuracy and scientific communication,” says Hernández Escayola. “Since we’re not going to be able to change young people’s habits, how can we get the information from a scientific paper into the format they’re going to consume?”
This doesn’t mean responsible content has to be dull. Maldita.es adapts its factchecking for TikTok by using the platform’s language, formats and trends without abandoning its journalistic standards. Its goal is to simplify complex information without oversimplifying the evidence.
Producers can take a similar approach by involving technical, scientific, regulatory and communications teams at the beginning of a campaign. Waiting until an eye-catching claim is ready for approval before involving specialists creates unnecessary risk. Skilled creators know how to hold an audience’s attention, while subject experts can recognise where simplification starts to distort the facts.
Hernández Escayola sees scope for scientists and credible influencers to work more closely, combining expertise in food, health and science with the ability to reach audiences online. Such partnerships could help accurate information adopt the language and pace of social media without sacrificing evidence, context or transparency.
Effective communication should explain what’s known, what remains uncertain and what the evidence means in practical terms. Brands also need to disclose commercial relationships prominently and link claims to accessible evidence rather than relying on vague references to ‘studies’, ‘research’ or unnamed experts.
Companies would also benefit from establishing a response process before damaging claims gain momentum. Monitoring emerging narratives, appointing authorised spokespeople and preparing experts to respond in accessible language can reduce the time between a false allegation appearing and an evidence-based correction being published.
Dr Betty Chang, research area lead at the European Food Information Council in Brussels, says regulation is already available in many areas but isn’t always applied effectively.
“There is existing regulation. It’s just not very well enforced,” she says, pointing to undisclosed sponsorship as one example. “A lot of content that’s sponsored wasn’t labelled as sponsored, even though influencers are not allowed to do this.”
Dr Chang believes platforms have a role in flagging misinformation and making commercial relationships clearer. Trust certifications for reliable sources or a type of information labelling could also help consumers judge what they encounter online.
“The bottom line we have to consider is how to make it so that people are not incentivised to spread misinformation,” she says. “A lot of the interest behind it is commercial, whether it’s from industry or from influencers. They are making money off this.”
Ultimately, a food brand that borrows the techniques of misinformation may secure attention in the short term, but it also gives consumers another reason to distrust the industry’s claims.
Every use of ‘natural’, ‘detoxifying’, ‘toxin-free’ or a similarly loaded term can now be clipped, challenged and shared far beyond its original context. Staying within the law is only the starting point – the claim also needs to be transparent, properly sourced and robust enough to survive public scrutiny.
