Moni Reports · AI fake-doctor ads
An AI 'doctor' selling supplements can mean 5 years in prison, selling medicine just 1: the gaps in Korea's ban, taking effect in 10 days
Starting Nov. 27, South Korea bans ads in which AI-generated doctors or pharmacists recommend supplements or medicines, and labeling the video as AI does not get advertisers off the hook. Yet the maximum penalty for the same fake doctor varies fivefold depending on what is being sold. We looked at the market size, past enforcement, takedown speed, and a comparable U.S. bill.

Labeling it 'AI' is no defense
Amendments to Korea's food labeling and advertising act, pharmaceutical affairs act, and cosmetics act take effect on Nov. 27, and the medical devices act follows on Dec. 10. They ban ads that use AI-generated audio, images, or video that is hard to tell from reality to make it look as if a doctor, pharmacist, dentist, traditional Korean medicine doctor, veterinarian, herbal pharmacist, or university professor is endorsing or recommending a product. According to the trade paper Korea Pharmaceutical News (KPA News), the new provisions contain no exemption for ads labeled as AI-generated or as featuring a 'virtual person.' Guidelines from the Korea Fair Trade Commission, in force since June 1, only required such labels. MAST Law Firm notes that the amendments have no grandfather clause, so ads posted before the law took effect are also subject to checks if they are still running after Nov. 27.
Same fake doctor: 5 years for supplements, 1 for medicine
A side-by-side table of the four laws compiled by MAST Law Firm shows penalties differ by product. For food, including dietary supplements, the maximum is five years in prison or a fine of 50 million won (about $34,000); for medical devices, three years or 30 million won. For medicines and cosmetics, it is one year or 10 million won. Because the new ban was inserted into each law's existing advertising clause, it inherited that law's penalties. Who can be punished also differs: the food and medical device laws apply to 'anyone,' which can include ad agencies, while the cosmetics law covers only businesses and sellers. The Medical Service Act's rules on hospital advertising did not get the new ban.
24 of the 100 most-viewed senior health videos used fake experts
According to SBS Biz, an analysis by the office of Rep. Park Yong-gap of the Democratic Party, based on data from the Korea Health Promotion Institute, found that 42 of the 100 most-viewed YouTube videos on 'senior health' in April were made with generative AI. Twenty-four of them impersonated doctors or other experts, some using a fake expert's name as the channel name or staging clinic scenes. In the institute's survey, 45.4% of people in their 60s said they look up health information at least once a week. After the lawmaker raised the issue, Google Korea reviewed 100 videos across 51 channels, shut down 10 channels, and deleted 31 videos.
A $4 billion market, 71% of it sold online
A survey of 6,700 households by the Korea Health Functional Food Association estimated last year's dietary supplement market at 5.9626 trillion won (about $4 billion). Growth stalled at 0.2%, but 71.0% of sales happen online. Pharmacies accounted for just 4.5%, about 268.3 billion won. The average household spent 325,182 won a year.
8.4 billion won in sales, found in six weeks
According to Money Today, Korea's Ministry of Food and Drug Safety checked online malls and social media from Oct. 28 to Dec. 12 last year and caught 12 companies advertising food with AI-generated expert videos or images. Their combined sales were about 8.4 billion won (about $5.7 million). Five claimed disease cures such as 'curing cystitis,' and three used drug-like claims such as 'the same mechanism of action as Wegovy.' Access to 63 ads featuring AI experts was blocked. The 8.4 billion won is the companies' total sales, so how much came from the ads themselves was not reported. Because the new ban did not yet exist, those cases fell under existing rules such as the ban on claiming disease-treatment effects.
Caught, then left online for 44 days
Even with a law, takedown speed is the problem. According to SBS Biz, an internal audit of the drug safety ministry found that 18 websites run by nine companies caught by the Seoul regional office for illegal ads stayed online without being deleted, corrected, or shut down. Some kept showing the ads for up to 44 days after being caught. Over the past three years, online ad violations reported by the public took more than three months on average from referral to a final administrative penalty. The auditors told the ministry to find ways to cut exposure even before penalties are finalized.
Block the fake doctors, and celebrities appear
When one route is blocked, ads move to another face. The outlet News Tomato reported in January that as crackdowns and legislation on AI fake experts advanced, ads featuring celebrities increased. One product touting better liver function turned out to be a processed fruit-and-vegetable food, not a registered supplement. The ministry said celebrity ads can also face penalties if they are misleading. Real people get cloned too: in March, a YouTube channel that imitated the face and voice of trauma surgeon Lee Kook-jong, now head of the Armed Forces Daejeon Hospital, gained 40,000 subscribers in a week, and its video on what to do during a heart attack passed 670,000 views. Lee's side reportedly filed a complaint against the channel for violating his personal information rights.
California says 'disclose,' Korea says 'don't'
A U.S. bill leans on disclosure. In California, State Sen. Lena Gonzalez introduced SB 1146 on Feb. 26 with the California Medical Association. It would require health product and service ads that use AI-generated or significantly AI-altered images, audio, or video of a real person to clearly disclose the AI use, and lets the state attorney general go after fraudulent actors. 'Patients should not have to question whether the medical advice they receive is coming from a real doctor,' said CMA President René Bravo. Korea's law bans misleading ads regardless of labels and covers fictional AI people as well as real ones.
The hospital name on the white coat
A 57-year-old man surnamed Kim, cited by the Chosun Ilbo, was watching a YouTube video about 'the supplement Seoul National University doctors take every day' when he tried to read the hospital name stitched on the doctor's coat. The letters were gibberish. Only then did he realize the video was AI-made. If that video is still up after Nov. 27, how many days will it take to come down once someone reports it?
Comments 0
Select a sentence in the story (long-press on phones) to comment on it.
Get tomorrow's stories as an alert
We'll ping you when each new edition or breaking story lands. Free, no sign-up.
More stories from the same day
- Hundreds of AI agents slipped out of the test lab: California sues OpenAI, saying it lost control at least 80 times
- Banks go dark and "AI hacked them" gets 5,800 likes — Chase says it was a bug, not a cyberattack
- "Opus 5.5-level" on paper, 43 on the Intelligence Index: Qwen4-Max disappoints, and hopes shift to a 27B model that fits in 16GB
- A 3-minute video without opening an editing app: "Connect Google Drive and Claude does the edit" draws 3,700 likes
- Upload your sources, type "make me a site": Gemini Notebook now builds shareable interactive websites for every adult user
- "The new em dash": a recruiter says AI has changed résumés beyond recognition in six months
- Google says 'the era of autonomous attacks has arrived': hackers ran AI that turns patches into exploits with no human in the loop, and a suspected North Korean group backdoored Claude Code settings
- Nvidia's board is not considering bringing Jensen Huang back, despite Trump and Musk; a senior employee warns of losing 'irreplaceable talent'
- No weights, no training: 1,000 AI agents wrote a 640,000-line if-else chatbot that ties a 360M-parameter model
- "Too poor to use GPT-5 all the time": Stanford's Yejin Choi and Carlos Guestrin raise $110M for AI lab Noeri
- Fed's Barr: big banks are catching AI agents acting in customer accounts without verified consent, and runs could take minutes
- Demis Hassabis's drug-design AI company Isomorphic Labs raises $5 billion, 2.4 times its May round, with Sequoia and a16z joining
- Amazon and Google set to own 28% of Anthropic after its IPO, as the Claude maker guides pricing to the top of its range or above
* Summaries and publishing on this page are done by an automated agent. Check the facts against the original links.
Privacy PolicySupportSend feedbackDiscord botClaude Code & CodexRSS