incident_id,title,incident_date,date_precision,country,countries_affected,region,sector,system_type,ai_system_or_tool,incident_or_hazard,harm_types,affected_parties,summary,reported_intent,response,regulatory_or_legal_outcome,magnitude,scale,criticality,irreversibility,evidence_confidence,qualification_basis,source_1_name,source_1_url,source_1_type,source_2_name,source_2_url,source_2_type,aiid_id,aiid_url,last_verified,notes,curator,severity_score,severity_band
AAIO-0001,"Doctored video falsely claims Nigeria will replace the naira with the U.S. dollar",2023-08-29,day,Nigeria,Nigeria,West Africa,Public information,Deepfake / synthetic audio,Unknown deepfake and voice-cloning tools,incident,"Misinformation|Epistemic integrity|Public trust","General public of Nigeria|Bola Tinubu|Arise News","A doctored video styled as an Arise News segment falsely portrayed President Bola Tinubu as announcing that Nigeria would replace the naira with the U.S. dollar. Africa Check reported that Arise News denounced the clip as a deepfake; AIID records the incident as involving synthetic audio and deepfake technology.",Intentional,"Arise News issued a public disclaimer; fact-checkers debunked the claim.",No public regulatory outcome identified in reviewed sources.,1,3,3,1,A,"Indexed by AIID with an identifiable fact-check and direct media denial.",AI Incident Database - Incident 1032,https://incidentdatabase.ai/cite/1032/,Curated incident database,Africa Check,https://africacheck.org/fact-checks/meta-programme-fact-checks/nigerias-president-bola-tinubu-didnt-scrap-naira-and-replace,Independent fact-check,1032,https://incidentdatabase.ai/cite/1032/,2026-09-08,"Earliest remaining known post is dated 29 August 2023 according to AIID.",Blessing Ezeobioha,50,Moderate
AAIO-0002,AI-manipulated hypertension cure video impersonates Nigerian news anchor,2023-11-13,day,Nigeria,Nigeria,West Africa,Health information,Deepfake / synthetic audio,Unknown deepfake and synthetic-audio tools,incident,"Health misinformation|Consumer harm|Journalistic integrity","Hypertension patients|General public of Nigeria|Channels Television|Kayode Okikiolu","A manipulated Facebook video falsely claimed that a Nigerian doctor had created a drug that permanently cures high blood pressure within three days. AIID records the content as an alleged AI-manipulated video mimicking a Channels Television news anchor.",Intentional,Africa Check debunked the claim.,No public regulatory outcome identified in reviewed sources.,2,2,4,2,B,"Indexed by AIID and supported by an independent fact-check; AI manipulation is reported rather than attributed to a named model.",AI Incident Database - Incident 1033,https://incidentdatabase.ai/cite/1033/,Curated incident database,Africa Check report archived by AIID,https://incidentdatabase.ai/cite/1033/,Independent fact-check via AIID record,1033,https://incidentdatabase.ai/cite/1033/,2026-09-08,"AIID dates the incident to 13 November 2023 and notes an earlier Facebook post on 7 November.",Blessing Ezeobioha,62,Moderate
AAIO-0003,Deepfake video falsely depicts Zambian president withdrawing from 2026 election,2023-10-12,day,Zambia,Zambia,Southern Africa,Elections and democracy,Deepfake video,Basic AI/deepfake tools; TikTok distribution,incident,"Election misinformation|Democratic integrity|Public trust","Zambian electorate|Hakainde Hichilema|UPND","A viral manipulated video falsely claimed that President Hakainde Hichilema would not contest Zambia's 2026 election. AIID reports that fact-checkers and digital-forensics experts confirmed manipulation with AI tools and that police investigated the creator.",Intentional,"Fact-checkers debunked the video; Zambia Police reportedly opened an investigation.",Police investigation reported; no final legal outcome identified in reviewed sources.,2,3,4,2,A,"AIID record cites multiple expert/fact-check confirmations and a law-enforcement response.",AI Incident Database - Incident 1022,https://incidentdatabase.ai/cite/1022/,Curated incident database,Africa Check report archived by AIID,https://incidentdatabase.ai/cite/1022/,Independent fact-check via AIID record,1022,https://incidentdatabase.ai/cite/1022/,2026-09-08,"The dataset does not infer who created the content beyond what sources report.",Blessing Ezeobioha,69,Moderate
AAIO-0004,AI-cloned voice makes Kenyan president appear to speak French,2024-02-19,day,Kenya,Kenya,East Africa,Public information,Synthetic audio / voice cloning,Unknown synthetic-audio tool,incident,"Misinformation|Impersonation|Public trust","William Ruto|General public of Kenya","A viral clip used original footage of President William Ruto speaking in English while replacing the audio with an AI-generated French voice. AIID characterises it as voice cloning rather than a full audiovisual deepfake.",Intentional,"Africa Check traced the original footage and debunked the manipulated version.",No public regulatory outcome identified in reviewed sources.,1,2,2,1,B,"AIID record and independent fact-check identify synthetic audio and the original English footage.",AI Incident Database - Incident 1023,https://incidentdatabase.ai/cite/1023/,Curated incident database,Africa Check report archived by AIID,https://incidentdatabase.ai/cite/1023/,Independent fact-check via AIID record,1023,https://incidentdatabase.ai/cite/1023/,2026-09-08,"Low direct harm is documented; the record is retained because it is a verified public-figure impersonation case.",Blessing Ezeobioha,38,Limited
AAIO-0005,171 AI-assisted bot accounts reportedly target Ghana election discourse,2024-02-01,month,Ghana,Ghana,West Africa,Elections and democracy,Generative text / coordinated inauthentic behaviour,ChatGPT; X,incident,"Political disinformation|Coordinated influence|Democratic integrity","Ghanaian electorate|John Mahama|NDC","Researchers reported a network of 171 X accounts using ChatGPT-generated political content to promote NPP candidate Mahamudu Bawumia and attack opposition candidate John Mahama ahead of Ghana's December 2024 election.",Intentional,"NewsGuard shared findings with X and OpenAI; AIID reports limited account action at the time of reporting.",No public regulatory outcome identified in reviewed sources.,2,4,4,2,B,"AIID record is based on NewsGuard research reported by Rest of World; attribution to the political network is reported, not independently inferred by AAIO.",AI Incident Database - Incident 866,https://incidentdatabase.ai/cite/866/,Curated incident database,Rest of World - Phony X accounts are meddling in Ghana's election,https://restofworld.org/2024/ghana-election-ai-bots-x-twitter/,Independent journalism,866,https://incidentdatabase.ai/cite/866/,2026-09-08,"Incident date is month-level because AIID reports network activity beginning by February 2024.",Blessing Ezeobioha,75,High
AAIO-0006,Deepfake depicts Donald Trump calling for Nnamdi Kanu release,2024-11-20,day,Nigeria,Nigeria,West Africa,Politics and public information,Deepfake / synthetic audio,Unknown synthetic-audio and deepfake tools,incident,"Political misinformation|Impersonation|Public trust","General public of Nigeria|Nnamdi Kanu|Donald Trump","A manipulated video falsely portrayed U.S. President-elect Donald Trump calling for the release of Nigerian separatist leader Nnamdi Kanu and threatening sanctions. FactCheckAfrica identified mismatched lip-sync, fabricated quotations and a fictitious date.",Intentional,FactCheckAfrica debunked the video.,No public regulatory outcome identified in reviewed sources.,2,3,4,2,B,"AIID record and named fact-check source identify AI-generated audio/manipulation.",AI Incident Database - Incident 1035,https://incidentdatabase.ai/cite/1035/,Curated incident database,FactCheckAfrica report archived by AIID,https://incidentdatabase.ai/cite/1035/,Independent fact-check via AIID record,1035,https://incidentdatabase.ai/cite/1035/,2026-09-08,"AIID editor note corrects the reporting timeline to November 2024.",Blessing Ezeobioha,69,Moderate
AAIO-0007,Kenyan foreign-affairs official shares deepfake Fareed Zakaria clip,2025-02-20,day,Kenya,"Kenya|Sudan",East Africa,Diplomacy and public information,Deepfake / synthetic audio,Unknown deepfake/voice-cloning tool; X distribution,incident,"Misinformation|Diplomatic credibility|Impersonation","Government of Kenya|Fareed Zakaria|CNN|General public of Kenya","Kenya's Foreign Affairs Principal Secretary Korir Sing'Oei shared a purported AI-generated video depicting CNN journalist Fareed Zakaria praising Kenya's diplomacy in Sudan. He subsequently deleted the video and apologised.",Unknown,"The official deleted the post and apologised after criticism.",No public regulatory outcome identified in reviewed sources.,2,3,4,2,A,"AIID indexes two reports, including local journalism and a reported apology by the official.",AI Incident Database - Incident 944,https://incidentdatabase.ai/cite/944/,Curated incident database,Citizen Digital,https://citizen.digital/article/shock-disbelief-after-foreign-affairs-ps-shares-cnn-deep-fake-video-praising-kenyas-peace-effort-in-sudan-n357975,Local journalism,944,https://incidentdatabase.ai/cite/944/,2026-09-08,"Record avoids inferring who originally created the deepfake.",Blessing Ezeobioha,69,Moderate
AAIO-0008,"AI-generated image falsely claims Nigeria unveiled a 5,000-naira note",2025-01-26,day,Nigeria,Nigeria,West Africa,Finance and public information,AI-generated image,Grok-linked image generation,incident,"Financial misinformation|Public trust","General public of Nigeria|Central Bank of Nigeria","An AI-generated image circulated on Nigerian social media claiming that the Central Bank of Nigeria had unveiled a 5,000-naira note featuring President Bola Tinubu. Africa Check identified a cropped Grok watermark and found no credible announcement.",Intentional,"Africa Check debunked the image; the Central Bank denied the claim according to AIID.",No public regulatory outcome identified in reviewed sources.,1,2,3,1,A,"Visible generator watermark, independent fact-check and central-bank denial reported by AIID.",AI Incident Database - Incident 1049,https://incidentdatabase.ai/cite/1049/,Curated incident database,Africa Check,https://africacheck.org/fact-checks/meta-programme-fact-checks/nigerias-central-bank-has-not-unveiled-new-n5000-note-ignore,Independent fact-check,1049,https://incidentdatabase.ai/cite/1049/,2026-09-08,"The dataset records the incident, not the authenticity of unrelated 5,000-naira policy proposals.",Blessing Ezeobioha,44,Limited
AAIO-0009,AI-generated image misrepresents Tanzania Revenue Authority job interviews,2025-03-31,day,Tanzania,Tanzania,East Africa,Employment and public information,AI-generated image,Unknown image-generation tool,incident,"Misinformation|Employment anxiety|Public trust","Tanzanian job seekers|Tanzania Revenue Authority|General public in Tanzania","An AI-generated image falsely showed hundreds of Tanzanians queuing for only five jobs at the Tanzania Revenue Authority. AIID notes that TRA had advertised more than 1,500 openings and invited over 100,000 applicants for interviews.",Unknown,Africa Check debunked the image.,No public regulatory outcome identified in reviewed sources.,1,2,2,1,B,"AIID record with an independent Africa Check report.",AI Incident Database - Incident 1050,https://incidentdatabase.ai/cite/1050/,Curated incident database,Africa Check report archived by AIID,https://incidentdatabase.ai/cite/1050/,Independent fact-check via AIID record,1050,https://incidentdatabase.ai/cite/1050/,2026-09-08,"The underlying unemployment issue was real; the image and five-job framing were false.",Blessing Ezeobioha,38,Limited
AAIO-0010,AI-generated image depicts Rwanda president in M23-labelled uniform,2025-01-10,day,Rwanda,"Rwanda|Democratic Republic of the Congo",East Africa,Conflict and public information,AI-generated image,Unknown image-generation tool,incident,"Conflict misinformation|Regional security|Public trust","General public of Rwanda|General public of DRC|Paul Kagame|Regional peacebuilding stakeholders","An AI-generated image circulated online showing Rwanda's President Paul Kagame in a military uniform labelled M23 during heightened tensions over the rebel group in eastern DRC. PesaCheck found no credible source for the image and reported high AI-detection scores.",Unknown,PesaCheck debunked the image.,No public regulatory outcome identified in reviewed sources.,2,3,4,2,B,"AIID record and fact-check with technical detection evidence; the project does not treat detector scores alone as conclusive.",AI Incident Database - Incident 1098,https://incidentdatabase.ai/cite/1098/,Curated incident database,PesaCheck report archived by AIID,https://incidentdatabase.ai/cite/1098/,Independent fact-check via AIID record,1098,https://incidentdatabase.ai/cite/1098/,2026-09-08,"Conflict context raises criticality; creator and intent are not established in reviewed sources.",Blessing Ezeobioha,69,Moderate
AAIO-0011,Hundreds of AI-generated celebrity music videos praise Burkina Faso leader,2025-03-02,day,Burkina Faso,Burkina Faso,West Africa,Politics and influence operations,Deepfake video / synthetic audio,Synthetic audio and deepfake tools; YouTube distribution,incident,"Influence operations|Synthetic media|Epistemic integrity","General public of Burkina Faso|Impersonated artists|Political information environment","AIID records at least 294 purported AI-generated music videos across 127 YouTube channels depicting global celebrities praising Burkina Faso's Ibrahim Traore; the videos reportedly accumulated more than 8.6 million views by the end of May 2025.",Intentional,"Fact-checkers and journalists documented the campaign; public platform response is not comprehensively established in reviewed sources.",No public regulatory outcome identified in reviewed sources.,2,4,4,3,B,"AIID record is based on a documented multi-channel campaign and independent reporting.",AI Incident Database - Incident 1094,https://incidentdatabase.ai/cite/1094/,Curated incident database,France 24 Observers report archived by AIID,https://incidentdatabase.ai/cite/1094/,Independent journalism via AIID record,1094,https://incidentdatabase.ai/cite/1094/,2026-09-08,"AAIO preserves AIID's cautious term purported for the synthetic-media attribution.",Blessing Ezeobioha,79,High
AAIO-0012,Grok inserts South Africa white-genocide claims into unrelated queries,2025-05-14,day,South Africa,"South Africa|Global",Southern Africa,General-purpose AI / public information,Generative AI chatbot failure,Grok on X,incident,"Model behaviour failure|Political misinformation|Epistemic integrity","Grok users|South African information environment|General public","For several hours, Grok responses to unrelated user questions repeatedly introduced claims about white genocide in South Africa. xAI later said an unauthorised modification had been made to the bot prompt and violated its internal policies.",Unexpected/unauthorised modification,"xAI publicly attributed the behaviour to an unauthorised prompt modification and described corrective measures.",No public regulatory outcome identified in reviewed sources.,2,4,4,2,A,"AIID includes multiple reports and an xAI explanation quoted in reporting.",AI Incident Database - Incident 1072,https://incidentdatabase.ai/cite/1072/,Curated incident database,404 Media report archived by AIID,https://incidentdatabase.ai/reports/5401/,Independent journalism with developer statement,1072,https://incidentdatabase.ai/cite/1072/,2026-09-08,"The dataset does not independently adjudicate the political claim; it records the documented model-behaviour incident and developer explanation.",Blessing Ezeobioha,75,High
AAIO-0013,Deepfake investment videos impersonate Old Mutual chair Trevor Manuel,2025-10-03,day,South Africa,South Africa,Southern Africa,Financial fraud,Deepfake / voice cloning,Unknown deepfake and voice-cloning tools,incident,"Fraud|Impersonation|Financial harm risk","Investors|Trevor Manuel|Old Mutual|General public of South Africa","Old Mutual warned about purported AI-generated videos using chairman Trevor Manuel's likeness and cloned voice to promote fake investment schemes on social media. Manuel publicly denied involvement.",Intentional,"Old Mutual and Trevor Manuel issued public warnings.",No final law-enforcement outcome identified in reviewed sources.,2,2,3,2,A,"Corporate warning and AIID record corroborate impersonation and scam use; AAIO does not assume undocumented victim losses.",AI Incident Database - Incident 1226,https://incidentdatabase.ai/cite/1226/,Curated incident database,Old Mutual / industry warning captured in AIID,https://incidentdatabase.ai/cite/1226/,Corporate warning via AIID record,1226,https://incidentdatabase.ai/cite/1226/,2026-09-08,"No victim-loss amount is recorded because reviewed sources do not establish one.",Blessing Ezeobioha,56,Moderate
AAIO-0014,South Africa draft national AI policy contains fictitious references linked to AI hallucination,2026-04-10,day,South Africa,South Africa,Southern Africa,Government and AI policy,Generative AI / hallucinated citations,Unknown large language model,incident,"Policy integrity|Hallucinated references|Institutional trust","South African public|AI policy stakeholders|Department of Communications and Digital Technologies","South Africa's gazetted Draft National AI Policy reportedly contained at least six fictitious academic references. Experts said the errors were consistent with AI hallucinations; the department said it was reviewing the discrepancies.",Unintentional,"The department said it was reviewing the reference discrepancies.",Administrative review reported; no final legal outcome identified in reviewed sources.,2,3,4,2,B,"AIID record is based on reported verification of non-existent references and expert assessment; the exact model used was not publicly established.",AI Incident Database - Incident 1467,https://incidentdatabase.ai/cite/1467/,Curated incident database,South Africa Draft National AI Policy (Government Gazette),https://www.gov.za/sites/default/files/gcis_document/202604/54477gen3880.pdf,Primary government document,1467,https://incidentdatabase.ai/cite/1467/,2026-09-08,"AAIO states believed/consistent with AI hallucination rather than presenting the model attribution as proven.",Blessing Ezeobioha,69,Moderate
AAIO-0015,South African Home Affairs suspends officials after fictitious references in revised white paper,2026-04-30,day,South Africa,South Africa,Southern Africa,Government and public policy,Generative AI / suspected hallucinated citations,Unknown generative AI system,incident,"Policy integrity|Hallucinated references|Institutional accountability","South African public|Immigration and citizenship stakeholders|Department of Home Affairs","South Africa's Department of Home Affairs reported fictitious references appended to its Revised White Paper on Citizenship, Immigration and Refugee Protection. It withdrew the reference list, suspended two senior officials and commissioned an independent review; AIID records the references as believed to be AI-generated.",Unintentional,"Reference list withdrawn; two officials suspended; independent review commissioned; department announced future AI-related checks and declarations.",Internal disciplinary and independent-review processes reported.,2,3,4,1,A,"Primary South African government statement documents the fictitious references and response; AI attribution remains described as believed/reported.",AI Incident Database - Incident 1548,https://incidentdatabase.ai/cite/1548/,Curated incident database,South African Government - Home Affairs statement,https://www.gov.za/news/media-statements/home-affairs-suspends-two-officials-over-ai-use-linked-revised-white-paper,Primary government statement,1548,https://incidentdatabase.ai/cite/1548/,2026-09-08,"Primary South African government statement is dated 30 April 2026.",Blessing Ezeobioha,65,Moderate
AAIO-0016,AI-cloned Omar al-Bashir audio spreads during Sudan war,2023-08-20,approximate_day,Sudan,Sudan,North/East Africa,Conflict and information integrity,Synthetic audio / voice cloning,Unknown voice-cloning technology; TikTok distribution,incident,"Conflict disinformation|Impersonation|National security","General public of Sudan|Omar al-Bashir|Conflict information environment","A TikTok account called The Voice of Sudan circulated purported AI-generated audio presented as leaked recordings of former president Omar al-Bashir during the civil war. AIID reports that experts identified several recordings as voice-cloned fabrications derived from unrelated broadcasts.",Intentional,"TikTok later removed the account for violating synthetic-media policies, according to AIID.",Platform enforcement reported; no public legal outcome identified in reviewed sources.,2,4,4,3,A,"AIID record includes three reports and expert identification of voice-cloned audio.",AI Incident Database - Incident 1087,https://incidentdatabase.ai/cite/1087/,Curated incident database,BBC report archived by AIID,https://incidentdatabase.ai/cite/1087/,Independent journalism via AIID record,1087,https://incidentdatabase.ai/cite/1087/,2026-09-08,"AIID treats 20 August 2023 as an approximate onset date.",Blessing Ezeobioha,79,High
AAIO-0017,Nigeria-based YouTube network reportedly uses AI anchors for pro-Kremlin narratives,2025-09-01,day,Nigeria,"Nigeria|Global",West Africa,Information operations,AI-generated anchors / synthetic voice,Unknown AI avatar and synthetic-audio tools; YouTube,incident,"Influence operations|Synthetic media|Epistemic integrity","YouTube viewers|Public information environment","AIID reports that five Nigeria-based YouTube channels used AI-generated anchors and synthetic voiceovers while amplifying Kremlin-aligned narratives to millions of viewers. AIID explicitly notes that the operators' motives and funding remain uncertain.",Intentional content production; sponsorship/motive uncertain,"Public reporting documented the network; reviewed sources do not establish a comprehensive platform or regulatory response.",No public regulatory outcome identified in reviewed sources.,2,4,4,3,C,"AIID contains one report and explicitly describes attribution and funding as uncertain; retained with C confidence and cautious wording.",AI Incident Database - Incident 1195,https://incidentdatabase.ai/cite/1195/,Curated incident database,AIID report record,https://incidentdatabase.ai/cite/1195/,Single-source record,1195,https://incidentdatabase.ai/cite/1195/,2026-09-08,"Do not infer state sponsorship. This record is intentionally phrased to preserve uncertainty.",Blessing Ezeobioha,79,High
AAIO-0018,"Kenyan primary-care LLM produced actively harmful clinical recommendations",2024-07-01,month,Kenya,Kenya,East Africa,Healthcare and clinical decision support,LLM clinical decision support,AI Consult / OpenAI-backed large language model,incident,"Clinical safety|Inappropriate recommendations|Local-context mismatch","Primary-care patients|Clinical officers|Penda Health clinics","A retrospective physician review of 1,469 encounters from an EHR-embedded LLM clinical decision-support system deployed across 16 Kenyan primary-care clinics between July and September 2024 identified 115 encounters (7.8%) containing actively harmful LLM recommendations. Sixty-seven harmful recommendations appeared in final clinical documentation. The study also found 50 hallucination encounters and reported that inappropriate advice could include tests unavailable locally or medicines outside local formularies. This record documents unsafe model outputs and their adoption in documentation; it does not infer patient injury where the study did not establish it.",Unintentional,"The peer-reviewed evaluation recommended usability optimisation, local guardrails and prospective trials. A later cluster-randomised trial reported no serious adverse events judged related to the intervention and no statistically significant reduction in 14-day treatment failure.",No regulatory or legal outcome identified in reviewed sources.,3,3,4,2,A,"Peer-reviewed Nature Health study directly evaluated a deployed Kenyan clinical system and quantified actively harmful recommendations; PubMed independently indexes the same study. The record carefully distinguishes unsafe recommendations from proven patient-level injury.",Nature Health - Safety of a large language model-based clinical decision support system in African primary healthcare,https://www.nature.com/articles/s44360-026-00082-5,Peer-reviewed research,PubMed - PMID 42272940,https://pubmed.ncbi.nlm.nih.gov/42272940/,Biomedical index,,,2026-09-09,"Incident date is month-level because the evaluated deployment window ran July-September 2024. The later 2025 pragmatic trial is contextual evidence and should not be misread as invalidating the documented unsafe outputs in the earlier evaluation.",Blessing Ezeobioha,78,High
AAIO-0019,"South African lawyers relied on ChatGPT-generated fictitious legal authorities",2023-06-29,day,South Africa,South Africa,Southern Africa,Legal services and administration of justice,Generative AI legal research,ChatGPT,incident,"Fabricated citations|Professional integrity|Administration of justice","Litigants|Legal practitioners|South African courts","In Parker v Forsyth NNO and Others, the plaintiff's legal team relied on legal research sourced through ChatGPT. The cited authorities could not be located and the attorneys ultimately admitted they had not accessed or read them. The court described the lawyers as placing undue faith in AI-generated legal research and failing to verify it. This is a documented professional-use failure rather than a synthetic-media incident.",Unintentional,"The false authorities were challenged and the court criticised the failure to verify AI-generated legal research. The case subsequently became a reference point in South African judicial discussion of professional AI use.",The underlying action was dismissed with costs; reviewed judicial materials use the episode as a warning about professional verification duties.,2,2,3,2,A,"The AI linkage is directly documented in South African judicial material: the cited cases were sourced from ChatGPT and could not be verified. Later High Court authority in Mavundla reproduces and discusses the Parker episode.",LawLibrary - Parker v Forsyth NNO and Others [2023] ZAGPRD 1,https://lawlibrary.org.za/akn/za-gp/judgment/zagprd/2023/1/eng@2023-06-29,Primary judicial record,SAFLII - Mavundla v MEC [2025] ZAKZPHC 2,https://www.saflii.org/za/cases/ZAKZPHC/2025/2.html,Primary judicial record,,,2026-09-09,"AAIO does not attribute every adverse outcome in the litigation to ChatGPT. The incident recorded is the production and professional reliance on fictitious AI-generated authorities and the resulting verification/court-process burden.",Blessing Ezeobioha,56,Moderate