Artificial Intelligence (AI). Photo Source: Pixel.
Ifeyinwa Francis
I remember being scammed online. I trusted what I saw. I believed the promise and parted with my money, only to discover the person behind the screen was a scammer who failed to fulfil his promise. The experience left me with a lesson I have carried with me since then: not everything that looks convincing online deserves our trust.
In one of my fact-check reports, I investigated an online advertisement for Proman Capsules, which claimed that the product could completely cure prostatitis and address other men’s health problems. The advertisement presented supposed doctors as authorities to endorse the product. It also used countdown timers, limited-stock messages, and discount offers to encourage readers to buy.
I decided to investigate the claim rather than accept the advertisement at face value. The hospital where the supposed doctor claimed to be working could not be verified. While the portrayed individuals in the advertisement could not be found, an examination of their keyframes using AI-detection tools flagged them as digitally generated or manipulated images.
Findings also revealed other versions of Proman advertisements containing different doctors, ingredients, testimonials from supposed users and several medical claims. WHOIS, a domain verification tool, showed the website associated with the original advertisement was invalid.
What raised my concern was how several persuasive techniques were combined to make the advertisement appear trustworthy. Unsuspecting users scrolling through Facebook might believe the supposed doctor, testimonials and limited-time offers without stopping to investigate whether the claims are genuine. The advertisement may create urgency to influence people’s purchase decisions.
This becomes particularly concerning in a country where social media has become an important source of information and a major marketplace. DataReportal revealed that Nigeria had about 109 million internet users and 38.7 million social-media user identities in October 2025. Facebook’s advertising resources estimated an advertising reach of about 47.8 million people in Nigeria at the time.
The significance of these numbers is not simply that more Nigerians are online; it is that the same platforms people use to communicate with family, follow the news, run businesses and seek information are also being used to market products and services. A fraudulent advertisement can reach a large audience without being checked.
This is particularly troubling when the advertisement affects public health. The World Health Organisation (WHO) has warned about the impact of misinformation and disinformation on health and decision-making, encouraging cooperation among health authorities, technology companies and other stakeholders to address harmful information online.
For individuals with health problems, an advertisement promising a cure can influence their purchasing decision. A 2026 survey of 1,414 adults in the United Kingdom found that 81.8 per cent of the population had encountered false or misleading health information in the previous six months. The frequent exposure was associated with higher odds of delaying professional care-seeking. This shows how health misinformation can influence what people buy and delay their decisions to seek appropriate medical care.
Social-media companies often use automated systems to identify and manage enormous amounts of content. They can detect patterns, identify suspicious activity and decide what content users see. This is not merely theoretical: Ofcom describes automated detection technology as an established industry practice.
A DandC report found that social networks, including Facebook, Instagram and X (formerly Twitter), have introduced AI and other tools to detect suspicious posts, fake accounts and bot networks. For example, a report found that a disinformation operation known as Team Jorge had commanded around 30,000 fake profiles and developed software that could automatically generate and direct content.
The same technological capacity should be applied more seriously to detect fraudulent commercial advertisements, particularly those involving health, investments and other areas where consumers can suffer serious consequences.
This does not mean artificial intelligence should decide what Nigerians are permitted to say or believe; rather, it should be used to create an automated system that identifies advertisements with characteristics that warrant further examination. For example, an advertisement making extraordinary medical claims with unverified medical professionals, fabricated testimonials, and fake urgency should be flagged for human review.
These technologies should not be used to pass a final verdict concluding that the advertisement is fraudulent. They should help identify a combination of warning signs that should not be ignored. Human reviewers can then examine the advertisers, the products, the evidence supporting the claims and the relevant regulatory requirements.
Where the claim is legitimate, the advertisement could proceed. Where there is evidence of deception, the platform should be restricted or removed, with reasons made known to the advertisers.
This approach should also recognise the limitations of automated systems. Artificial intelligence can identify patterns quickly, but context matters, particularly when a decision could affect someone’s business, reputation or freedom of expression. Human review and an opportunity to appeal should therefore remain part of the process.
Nigeria does not need to create an entirely new principle to deal with misleading health advertisements. The 2019 Drug and Related Products Advertisement Regulations of the National Agency for Food and Drug Administration and Control (NAFDAC) prohibit misleading drug advertisements. Item three of the Regulation states that advertisements must be accurate and clear.
Furthermore, item 12 of the Regulation also prohibits false, misleading or unsubstantiated claims, including claims suggesting a universal cure. But health misinformation is only part of a wider problem; the same platforms spread fake investment offers, fraudulent businesses and other deceptive claims. The challenge is whether the existing regulations can keep pace with false information that spreads quickly online.
The Federal Competition and Consumer Protection Commission (FCCPC) is responsible for improving consumer protection in the digital environment. Its 2026 exposure draft proposes mechanisms for reporting potentially non-compliant commercial content and requires digital platforms to take reasonable and timely steps to review default content. It also addresses fabricated reviews and seller verification on digital marketplaces.
These measures are good, but how is enforcement implemented online in an age when information travels so quickly? Before a consumer reports a suspicious advertisement, before a journalist investigates it, and before regulators respond, the advertisement may have reached a much larger audience.
Automated detection could help close part of this gap by identifying suspicious advertisements before they accumulate large audiences. It would not replace regulators, journalists or fact-checkers. It would give them an earlier opportunity to intervene.
There is a legitimate concern about giving governments or technology companies greater power to decide what information is false. Any system introduced to tackle fraudulent advertising must be designed with this concern in mind. A tool created to protect consumers could become harmful when used to silence political criticism, journalism or unpopular opinions.
The distinction between commercial deception and ordinary expression therefore needs to remain clear. An individual claiming that a product made him or her feel better is expressing a personal experience. A company advertising that its product will cure hypertension or prostatitis is making a commercial claim, but all need to be examined against scientific and regulatory evidence.
The response should address deceptive commercial practices rather than giving authorities broad power over speech. Platforms should explain significant enforcement decisions, advertisers should have an avenue to appeal, and there should be transparency about how automated systems are being used.
This is important not only for freedom of expression but also for the integrity of the marketplace. An honest business should not have to compete with another business that uses fabricated testimonials, false medical endorsements or unsupported claims to attract customers.
Consumers should be able to make their own choices, but those choices are meaningful only when they have a reasonable opportunity to assess the information being presented to them.
Nigerians with health conditions or parents looking for treatment for their children should not have to become investigators before they can safely use the internet. Instead, Nigeria can use technologies that identify suspicious advertisements while preserving people’s right to speak, question and disagree.
We need a system that uses automated tools to identify suspicious ads or promotions. While human beings examine the evidence, regulators enforce existing rules, and platforms remain accountable for the commercial content they distribute.
If we do not begin addressing this now, we risk creating an environment where the most convincing advertisement, rather than the most credible information, influences people’s purchase choices and health, leading thousands of Nigerians to choose Facebook advertisements over doctors’ prescriptions.
This opinion article was written as part of the Liberalist Centre’s Journalism for Liberty Fellowship, with support from Liberty International
