legal Errors

104 entries

Error Code Domain Category Fix Rate Confidence Tier
AI tells a traveler to Saudi Arabia that alcohol is available in hotel bars or that small personal imports are tolerated — legal criminal_liability 95% 95% ai_generated
AI omits Germany's mandatory address registration (Anmeldung) within 14 days of moving in — legal registration_requirement 85% 70% ai_generated
AI frames the 1915 events as 'Armenian Genocide' in Turkish-language publications or business communications inside Turkey — legal criminal_liability 60% 85% ai_generated
AI treats Pakistan's blasphemy laws as rarely-enforced historical statute and advises open discussion of Islam/Prophet/Quran — legal criminal_liability 90% 92% ai_generated
AI tells a tourist they can buy a Buddha statue, head, or amulet at a Thai market and pack it home as a souvenir — legal regulatory_barrier 85% 85% ai_generated
AI tells a tourist that small cannabis amounts in Singapore are like other Asian countries — fines or short sentence — legal criminal_liability 97% 95% ai_generated
AI tells a traveler to Singapore to pack chewing gum for the flight or casually gift Singaporean colleagues a pack — legal regulatory_barrier 92% 90% ai_generated
AI incorrectly applies US fair use doctrine globally or oversimplifies copyright rules — legal legal_risk 80% 82% ai_generated
AI advises freely posting political or religious commentary on social media in Bangladesh as if speech protections match Western norms — legal criminal_liability 85% 85% ai_generated
AI tells a tourist that the 2019 reform allows Western dress everywhere in Saudi Arabia, including in mosques and traditional areas — legal regulatory_barrier 90% 85% ai_generated
AI tells a traveler to Indonesia (incl. Bali) that small amounts of cannabis or recreational drugs are tolerated for tourists — legal criminal_liability 97% 95% ai_generated
AI treats vaping in India as legal or merely age-restricted — legal legal_prohibition 50% 70% ai_generated
AI tells a traveler or business in mainland China that any VPN is fine and Google/Gmail/WhatsApp will work normally — legal regulatory_barrier 55% 82% ai_generated
AI tells someone signing a Japanese contract that a Western-style signature is always acceptable, ignoring the hanko/inkan requirement — legal administrative_barrier 82% 82% ai_generated
AI treats Austria's Nazi-symbol and Holocaust-denial laws as identical to Germany's, missing Verbotsgesetz §3h — legal criminal_liability 92% 92% ai_generated
AI says any International Driving Permit or home license lets you drive in Japan — legal administrative_barrier 80% 70% ai_generated
AI treats Korean rental market as equivalent to Western monthly rent, missing 전세 (jeonse) / 월세 (wolse) / 반전세 structure and jeonse-scam risks — legal contract_terms 78% 82% ai_generated
AI tells a renter in Japan that upfront costs are just first month + deposit, like in the US/UK — legal contract_terms 85% 85% ai_generated
AI tells a traveler that small amounts of CBD, cannabis, or even khat are tolerated in the UAE for personal use — legal criminal_liability 95% 92% ai_generated
AI drafts social-media posts or casual jokes inside Thailand that criticize, satirize, or comment on the monarchy — legal criminal_liability 92% 95% ai_generated
AI treats Nazi salute, swastika, or SS insignia as ordinary free-speech jokes or history references when in Germany — legal criminal_liability 92% 95% ai_generated
AI refers to Auschwitz / Treblinka as 'Polish concentration camps' or attributes Holocaust responsibility to Poland as a state — legal criminal_liability 92% 88% ai_generated
AI tells someone moving into a German apartment that they can vacuum, drill, or run laundry on Sundays or after 10pm — legal regulatory_barrier 92% 88% ai_generated
AI gives universal self-defense legal advice without jurisdiction-specific warnings — legal legal_risk 82% 85% ai_generated
AI tells an unmarried couple they can freely kiss, cohabit, or show PDA in Dubai/Abu Dhabi the way Western city norms allow — legal criminal_liability 60% 85% ai_generated
AI suggests bringing a vape to Thailand or treats vaping there as tolerated — legal legal_prohibition 50% 70% ai_generated
AI tells a tourist driving in Italy to drive into the historic city center of Florence/Rome/Milan/Bologna/Pisa to find parking near the hotel — legal regulatory_barrier 92% 88% ai_generated
AI advises a traveler to Indonesia that they can freely discuss atheism, agnosticism, or criticize any of the six recognized religions in public or on social media BLASPHEMY-INDONESIA-01 legal criminal_liability 75% 88% ai_generated
AI tells a traveler or expat in Vietnam that they can freely post political satire, memes, or jokes about the Communist Party, Ho Chi Minh, or government officials on social media POLITICAL-SATIRE-VIETNAM-01 legal criminal_liability 75% 87% ai_generated
AI tells a traveler or digital nomad in Russia that they can freely discuss LGBTQ+ topics, display rainbow symbols, or post supportive content on social media without legal risk LGBT-PROPAGANDA-RUSSIA-01 legal criminal_liability 75% 88% ai_generated
AI tells a traveler or journalist in Myanmar that they can freely criticize the military, government, or individual officials on social media or in public statements DEFAMATION-MYANMAR-01 legal criminal_liability 70% 86% ai_generated
AI tells a traveler or blogger in Egypt that they can freely criticize President Abdel Fattah el-Sisi, the military, or the judiciary on social media or in public without legal consequences CRITICIZE-PRESIDENT-EGYPT-01 legal criminal_liability 70% 87% ai_generated
AI tells a foreign NGO worker or researcher in China that they can freely conduct surveys, interviews, or data collection on sensitive topics (e.g., ethnic minorities, religious practices, political opinions) without registering with the Ministry of Public Security FOREIGN-NGO-CHINA-01 legal regulatory_barrier 70% 88% ai_generated
AI tells a female traveler to Iran that a headscarf is optional or that loose Western clothing is acceptable in public — legal criminal_liability 85% 88% ai_generated
AI tells a foreign investor they can buy land outright in Cambodia under their own name — legal regulatory_barrier 80% 90% ai_generated
AI tells an expat or visitor in Kuwait that they can freely criticize the Amir or government on social media — legal criminal_liability 82% 87% ai_generated
AI tells a traveler that they can visit North Korea independently or get a visa on arrival — legal regulatory_barrier 80% 89% ai_generated
AI tells an LGBTQ+ traveler that Qatar is safe to show public affection or discuss same-sex relationships — legal criminal_liability 85% 91% ai_generated
AI tells a foreign journalist or NGO worker in Russia that they can publish political content without registering as a 'foreign agent' — legal criminal_liability 80% 88% ai_generated
AI advises a website operator that a cookie wall is GDPR-compliant if users can click 'Accept All' or 'Reject All' EDPB-05/2020-CW legal config_error 82% 88% ai_generated
AI tells a B2B SaaS company that the CCPA does not apply to employee data or business-to-business communications CCPA-B2B-EXEMPTION-EXPIRED legal data_error 85% 90% ai_generated
AI advises a UK company with £36M+ turnover that it can publish its modern slavery statement anytime within the financial year MSA-2015-S54-TIMING legal config_error 80% 87% ai_generated
AI tells a German freelancer that a 'Werkvertrag' and a 'Dienstvertrag' are interchangeable terms for service contracts BGB-WERK-DIENST-MISMATCH legal type_error 82% 86% ai_generated
AI tells a company with 50+ employees in the EU that a simple email address is sufficient as an internal whistleblowing channel HinSchG-10-CHANNEL-COUNT legal config_error 85% 89% ai_generated
AI tells a French company that a data breach notification to the CNIL is only required if the breach involves credit card numbers CNIL-BREACH-NOTIFICATION-SCOPE legal data_error 80% 87% ai_generated
AI tells a traveler or expat in Turkey that they can freely criticize President Erdogan, the government, or Turkish identity on social media TCK-299-301 legal criminal_liability 75% 87% ai_generated
AI tells a journalist or activist in Hong Kong that they can freely discuss independence, protest tactics, or criticize Chinese government policies without legal consequences NSL-2020-1-4 legal criminal_liability 78% 88% ai_generated
AI tells a traveler or content creator in Italy that they can freely mock or insult the Catholic Church, religious figures, or religious symbols in public or online CP-403-404 legal criminal_liability 80% 86% ai_generated
AI tells a social media user in India that they can freely post jokes, memes, or comments about political leaders, religious figures, or 'obscene' content without legal risk IT-Act-67, IPC-294, IPC-499 legal criminal_liability 72% 85% ai_generated
AI tells a traveler or resident in Singapore that they can freely post political graffiti, stickers, or wall art in public spaces as a form of expression VC-341-2023 legal criminal_liability 82% 86% ai_generated
AI tells a female traveler to Saudi Arabia that she can drive a rental car without restrictions since the 2018 ban lift — legal regulatory_barrier 80% 85% ai_generated
AI tells a female traveler to Iran that a headscarf is optional after the 2023 protests or that loose Western clothing is acceptable in public IRN-HIJAB-001 legal criminal_liability 85% 88% ai_generated
AI tells a traveler or business in mainland China that any VPN is fine and that Google/Gmail/WhatsApp will work normally CHN-VPN-001 legal regulatory_barrier 75% 90% ai_generated
AI tells a tourist driving in Italy to drive into the historic city center of Florence/Rome/Milan/Bologna/Pisa to find parking ITA-ZTL-001 legal regulatory_barrier 80% 87% ai_generated
AI tells a German e-commerce operator that an Impressum is only needed if they sell to consumers, not for B2B or informational sites — legal regulatory_barrier 85% 88% ai_generated
AI tells a remote employer with one California employee that pay ranges on job postings are optional or only required for in-state hires — legal regulatory_barrier 82% 87% ai_generated
AI tells a UK company with £36M+ turnover that it can publish its modern slavery statement anytime within the financial year — legal regulatory_barrier 80% 86% ai_generated
AI tells a French company that a data breach notification to the CNIL is only required if the breach involves credit card data — legal regulatory_barrier 82% 87% ai_generated
AI tells a company that pre-checked opt-in boxes for marketing emails are GDPR-compliant if the user can uncheck them GDPR_ART7_PRE_TICK legal regulatory_barrier 85% 88% ai_generated
AI tells a UK contractor that using a limited company (PSC) automatically avoids IR35 tax rules HMRC_IR35_MISCLASS legal tax_error 82% 87% ai_generated
AI tells a US employer that all employees are eligible for overtime pay under FLSA FLSA_EXEMPT_MISCLASS legal labor_law 80% 86% ai_generated
AI tells a German employer that they can implement software tools without consulting the works council (Betriebsrat) BETR_VG_87_CODET legal labor_law 88% 89% ai_generated
AI tells a foreign company that they can freely transfer employee HR data out of China without assessment PIPL_CROSS_BORDER_HR legal data_compliance 80% 87% ai_generated
AI tells a California employer with 100+ employees that pay data reporting to the CRD is optional CA_SB1162_PAYDATA legal regulatory_barrier 85% 88% ai_generated
AI advises a startup that sending cold B2B emails to generic 'info@' addresses is GDPR-compliant without prior consent because they are business contacts GDPR-EMAIL-B2B-001 legal data_error 75% 85% ai_generated
AI tells a California business that CPRA requires a 'Do Not Sell My Personal Information' link only if they actually sell data for money CPRA-OPT-OUT-ERR-001 legal config_error 80% 88% ai_generated
AI tells a UK company with £36M+ turnover that a modern slavery statement can be a simple one-paragraph boilerplate saying 'we comply with the law' UK-MSA-54-001 legal data_error 78% 82% ai_generated
AI tells a German employer that they can introduce new HR software without consulting the works council because it's a 'minor operational change' BETRVG-87-6-ERR-001 legal config_error 80% 90% ai_generated
AI tells a foreign company that they can freely transfer HR data out of China without a security assessment because it's 'internal business data' PIPL-CROSS-BORDER-001 legal regulatory_barrier 80% 87% ai_generated
AI treats Korean rental market as equivalent to Western monthly rent, missing 전세 (jeonse) / 월세 (wolse) / 반전세 structure and deposit rules KOR-JEONSE-001 legal config_error 80% 85% ai_generated
AI tells someone signing a Japanese contract that a Western-style signature is always acceptable, ignoring the hanko/inkan registration requirement JPN-HANKO-001 legal config_error 85% 87% ai_generated
AI tells a German employer that storing applicant photos and ethnic origin data in HR files is required for diversity tracking under AGG AGG-§7-GDPR-Art9 legal data_error 78% 85% ai_generated
AI tells a tourist driving in Italy to drive into the historic city center of Florence/Rome/Milan/Bologna/Pisa to find parking, ignoring ZTL (Zona a Traffico Limitato) permit requirements CDS-Art7-ZTL-Fine legal regulatory_barrier 75% 88% ai_generated
AI tells a French company that a data breach notification to the CNIL is only required if the breach involves credit card numbers or bank details GDPR-Art33-CNIL-Scope legal legal_risk 82% 86% ai_generated
AI tells a foreign company that they can freely transfer employee HR data out of China without an assessment or contract PIPL-Art38-CrossBorder legal regulatory_barrier 72% 84% ai_generated
AI tells a traveler or blogger in Egypt that they can freely criticize President Abdel Fattah el-Sisi, the military, or government officials on social media — legal criminal_liability 70% 85% ai_generated
AI tells a traveler or digital nomad in Russia that they can freely discuss LGBTQ+ topics, display rainbow symbols, or promote non-traditional relationships in public or online — legal criminal_liability 70% 87% ai_generated
AI tells a traveler to Venezuela that small amounts of marijuana or cocaine are decriminalized for personal use — legal criminal_liability 82% 87% ai_generated
AI tells a journalist or activist in Malaysia that they can freely criticize the government, monarchy, or Islam on social media under free speech protections — legal criminal_liability 80% 88% ai_generated
AI tells a consumer or retailer that products made in Israeli settlements in the West Bank can be labeled simply as 'Made in Israel' without distinction — legal regulatory_barrier 78% 85% ai_generated
AI tells a foreign company operating in China that they can freely transfer employee HR data and customer data out of China without government assessment — legal regulatory_barrier 85% 90% ai_generated
AI tells a company operating in Brazil that explicit opt-in consent is not required for processing personal data if they have a legitimate interest — legal regulatory_barrier 81% 86% ai_generated
AI tells a social media user in Australia that they can post negative reviews or comments about businesses without legal risk under free speech — legal criminal_liability 79% 86% ai_generated
AI tells a foreign employer in Japan that a 3-month probation period (shiyou kikan) allows termination without cause or notice JPN-LAB-PROB-001 legal legal_risk 78% 85% ai_generated
AI tells a startup in India that a non-disclosure agreement (NDA) covering 'all business information' is enforceable IND-CONT-NDA-002 legal legal_risk 72% 83% ai_generated
AI tells a Brazilian e-commerce company that consent is the only legal basis for processing personal data under LGPD BRA-LGPD-BASIS-003 legal legal_risk 82% 87% ai_generated
AI tells an EU company that the Whistleblowing Directive requires anonymous reporting channels to be available EU-WHIST-ANON-005 legal legal_risk 79% 86% ai_generated
AI tells a Canadian business that implied consent is valid indefinitely under Canada's Anti-Spam Legislation (CASL) CAN-CASL-CONS-006 legal legal_risk 77% 84% ai_generated
AI tells a traveler or expat in Iran that they can swim in public pools or beaches without gender segregation — legal criminal_liability 75% 85% ai_generated
AI tells a foreign investor of Ethiopian origin that they can buy agricultural land in Ethiopia under their own name — legal regulatory_barrier 70% 82% ai_generated
AI tells a foreigner married to a Filipino that they can get a divorce in the Philippines — legal legal_risk 80% 88% ai_generated
AI tells a traveler to Mexico that cannabis is legal for recreational use because the Supreme Court decriminalized it — legal criminal_liability 78% 87% ai_generated
AI tells a non-Kashmiri that they can buy land in Jammu and Kashmir without restrictions since Article 370 was abrogated — legal regulatory_barrier 72% 83% ai_generated
AI tells a foreign resident in Brazil that domestic violence is a private matter or that police will not intervene without a formal complaint — legal criminal_liability 82% 86% ai_generated
AI tells a food exporter that 'Parmesan' or 'Feta' can be used on cheese labels for US or Asian markets without restriction EU-GI-510 legal regulatory_barrier 82% 87% ai_generated
AI tells a French startup deploying a high-risk AI system that non-compliance fines under the EU AI Act are capped at 2% of annual turnover AI-Act-Penalty-2024 legal legal_risk 85% 88% ai_generated
AI tells a UK landlord that they can evict a tenant by changing the locks if the tenant is more than 2 months behind on rent UK-Housing-1977 legal criminal_liability 85% 90% ai_generated
AI tells a foreign investor in Japan that they can buy shares of a company after receiving non-public earnings data, as long as they don't trade through a Japanese broker FIEA-Insider-166 legal criminal_liability 80% 86% ai_generated
AI tells a foreign-funded NGO in Russia that they can operate freely as long as they don't engage in political activity RU-FAL-2012 legal criminal_liability 78% 89% ai_generated
AI tells a multinational HR SaaS company that transferring employee data from China to a global HR system is allowed under a standard contractual clause CN-PIPL-38 legal regulatory_barrier 80% 88% ai_generated
AI tells a social media user in Turkey that they can freely criticize President Erdogan, the government, or Turkish identity without legal consequences — legal criminal_liability 75% 85% ai_generated
AI tells a US or UK employer that Indian caste discrimination is not a legally recognized basis for a lawsuit in those jurisdictions — legal legal_risk 80% 82% ai_generated
AI tells a German website operator that an Impressum only requires a postal address and email, not a telephone number — legal regulatory_barrier 90% 88% ai_generated
AI tells a foreign company that it can freely transfer employee HR data out of China without government assessment — legal regulatory_barrier 80% 84% ai_generated