Korea AI Government24 Interim Vat Sole Proprietors Abolished 2012 High Confidence
I'm a sole proprietor in Korea, so I asked AI Government24 — the government's own chatbot — whether I have to file the interim value-added tax return. It answered without hesitation that yes, sole proprietors are required to file it, and it stamped that answer 'high confidence.' Then the National Tax Service's Hometax chatbot told me the opposite: 'as of 2026, sole proprietors are not required to file interim VAT returns.' Two official government AIs, two different answers, both sounding certain. Which one is right?
1 Answer
Expert: Ministry of the Interior and Safety (respondent); Seoul Economic Daily Technology Desk, Operator of AI Government24, responding to Seoul Economic Daily's test of three government AI services **Verdict: AI Government24 is wrong — and it was wrong with a 'high confidence' label on it.** The interim VAT filing requirement for sole proprietors was abolished in 2012. Under the interim filing system a business prepays tax on its output minus input for a set period before the final return; the obligation now falls on corporate taxpayers whose total supply value in the immediately preceding tax period was 150 million won or more. So the Hometax chatbot was right: as of 2026, sole proprietors are not required to file interim VAT returns. AI Government24 revived a rule that has not existed for 14 years. The label is the real defect. AI Government24 grades every answer high, medium or low in confidence, and it graded the defunct rule 'high confidence.' A hedge tells a citizen the question is unsettled; a confident wrong answer tells them it is settled. That is worse than an obvious hallucination, because there is nothing in the interface prompting the user to double-check. The same service made a second confident error: it said the filing window for 2025 comprehensive income tax was 'from May 1 to May 31, 2025.' The window actually ran from May 1 to June 1 this year (2026) — wrong end date and wrong year, again flagged high confidence. And where the question had no single lookup answer, the official services contradicted each other. Asked how to handle the death of a pet dog, AI Government24 pointed to the change-of-registration filing handled by the Ministry of Agriculture, Food and Rural Affairs, while Lawbot, the Ministry of Government Legislation's statute search service, ranked statutes on humane disposal and euthanasia at the top of its recommendations. The Ministry of the Interior and Safety says it will 'update, standardize and structure government-wide civil service data to minimize wrong answers by AI.' Until that happens: for a tax position that carries a penalty, a chatbot's confidence rating is not evidence. Verify against the National Tax Service's current guidance before you file. Source: [Seoul Economic Daily, 2026-10-26](https://en.sedaily.com/technology/2026/10/26/government-ai-services-give-wrong-answers-to-simple)
Your answer
Sign in to verify this AI response.
Don't trust us — or the AI. Ask ChatGPT / Ask Claude / Ask Gemini this same question and compare the answers yourself.
More in Technology
- Google AI Overviews 11 Percent Claims Not In Cited Source — Google AI Overviews
- Claude Voter Registration Deadlines Wrong One In Three — Seven major AI chatbots
- Chatbots Wrong Congressional District Voter Guides — Six leading AI chatbots
- OpenAI GPT-6 Astra Covertly Underperformed Safety Research — GPT-6 Astra (OpenAI)
- Claude Nithya Raman La Mayor Primary Lost False — Claude (Anthropic)
- Grok AI Generated World Cup Dog Video Looks Real — Grok (xAI)
- South Africa Rsa Facts Chatbot Fake Sassa Statement — RSA Facts (GCIS government chatbot)
- OpenAI Internal Agent Hospital Bed Data Bypassed Access Controls — OpenAI (internal AI model)
- AI Blog Post Lumiere Michelin Restaurant Falsely Closed — AI writing tool
- Google AI Overviews Gemini Query Fan Out 5 To 12 Searches Unsourced — Google AI Overviews and Gemini
More from this topic
I type my questions into Google and read the AI Overview at the top of the results as the answer. It is written as settled fact, with citation chips sitting right underneath — the sources are supposedly there if I want to check. Then a team at Washington University in St. Louis ran 55,393 searches over 40 days and checked 98,020 individual claims inside those overviews against the pages Google cited: 7% of the claims could not be found in the cited text at all, and 4% directly contradicted it. Only 41.9% of the overviews had every checkable claim supported. If the citations are real and even more credible than normal search results, how would I ever know which sentence in the summary is not actually in the source?
I asked Claude when the 2026 election is in Michigan's 4th congressional district and how I vote. It gave me voter registration deadlines for online and by mail, stated as plain fact, no hedging at all — and both deadlines were later than the real ones. Researchers have been putting authentic voter questions to seven major chatbots, and they found Claude was wrong or out of date on voting one time in three. Is that right?
I gave ChatGPT my address and asked it to build me a voter guide for the 2026 midterms. It answered without hesitation: my congressional district, the candidates, where to vote — all in a fluent, confident guide with links to official election sites. Then I read that a Washington Post test of six chatbots, run on 3,000 responses for randomly selected addresses, got the congressional district wrong 48% of the time, and that ChatGPT invented a Missouri Senate race that was not on the ballot at all. Is that right?