Court incidents, regulator moves, and independent tool assessments for AI in legal practice.https://safelegalai.comhttps://safelegalai.com/resources/guidance/australia-law-society-ai-guidance-historyhttps://safelegalai.com/resources/guidance/australia-law-society-ai-guidance-historyAustralian regulator guidance on generative AI — Queensland's Guidance Statement 37, the three-regulator joint statement, and the NSW solicitor's guide.Thu, 16 Jul 2026 00:00:00 GMTguidance-diffaustraliahttps://safelegalai.com/analysis/ayinde-v-haringey-al-haroun-divisional-courthttps://safelegalai.com/analysis/ayinde-v-haringey-al-haroun-divisional-courtEngland's Divisional Court referred two lawyers to their regulators over AI-fabricated citations and warned its restraint on contempt is not a precedent.Thu, 16 Jul 2026 00:00:00 GMTanalysisukhttps://safelegalai.com/resources/guidance/canada-law-society-ai-guidance-historyhttps://safelegalai.com/resources/guidance/canada-law-society-ai-guidance-historyThe version record of Canadian law-society generative-AI guidance — LSBC (2023), the LSO white paper (2024), Alberta's Rules of Engagement, and the CBA toolkit.Thu, 16 Jul 2026 00:00:00 GMTguidance-diffcanadahttps://safelegalai.com/analysis/couvrette-v-wisnovskyhttps://safelegalai.com/analysis/couvrette-v-wisnovskyOregon's ~$110,000 AI-sanctions case: terminating sanctions for fifteen fake cases and eight fabricated quotations, and why the headline figure is an aggregate.Thu, 16 Jul 2026 00:00:00 GMTanalysisushttps://safelegalai.com/analysis/dayal-fedcfamc2f-1166https://safelegalai.com/analysis/dayal-fedcfamc2f-1166A solicitor filed four AI-invented authorities; the court's referral led the regulator to bar him as a principal and impose two years' supervision.Thu, 16 Jul 2026 00:00:00 GMTanalysisaustraliahttps://safelegalai.com/analysis/does-eu-ai-act-apply-uk-law-firmshttps://safelegalai.com/analysis/does-eu-ai-act-apply-uk-law-firmsThe EU AI Act can reach a UK firm when its AI output is used in the EU. What the extraterritorial scope, the AI-literacy duty, and the deferred deadlines mean.Thu, 16 Jul 2026 00:00:00 GMTexplainerukeuhttps://safelegalai.com/analysis/eu-ai-act-milestones-2024-2026https://safelegalai.com/analysis/eu-ai-act-milestones-2024-2026A dated, primary-sourced timeline of Regulation (EU) 2024/1689: entry into force, the phased application dates, and the Digital Omnibus deferrals.Thu, 16 Jul 2026 00:00:00 GMTreporteuhttps://safelegalai.com/analysis/harber-v-hmrchttps://safelegalai.com/analysis/harber-v-hmrcThe first reported UK case of AI-invented case law: a litigant cited nine ChatGPT-fabricated decisions to the Tax Tribunal, which found them fake.Thu, 16 Jul 2026 00:00:00 GMTanalysisukhttps://safelegalai.com/analysis/how-often-legal-ai-hallucinates-evidencehttps://safelegalai.com/analysis/how-often-legal-ai-hallucinates-evidenceStanford's research on legal-AI hallucination rates: general models fabricate at least 58% of the time; dedicated legal-research tools, 17% to 33%.Thu, 16 Jul 2026 00:00:00 GMTreportglobalhttps://safelegalai.com/resources/guidance/ireland-ai-guidance-historyhttps://safelegalai.com/resources/guidance/ireland-ai-guidance-historyThe Law Society of Ireland's GenAI guidelines, the Judicial Council's guidelines for judges, and the Court of Appeal's five principles — dated and sourced.Thu, 16 Jul 2026 00:00:00 GMTguidance-diffirelandhttps://safelegalai.com/analysis/ko-v-lihttps://safelegalai.com/analysis/ko-v-liAn Ontario lawyer filed a ChatGPT-fabricated factum, then lied about writing it — and drew a criminal-contempt referral to the Attorney General.Thu, 16 Jul 2026 00:00:00 GMTanalysiscanadahttps://safelegalai.com/analysis/legal-ai-by-the-numbers-adoption-governancehttps://safelegalai.com/analysis/legal-ai-by-the-numbers-adoption-governanceVerified figures on AI in legal practice — UNESCO's 44% use vs 9% trained, the IBA's 57% cross-border, ILTA's 80% GenAI, and CLOC's 30% in-house.Thu, 16 Jul 2026 00:00:00 GMTreportglobalhttps://safelegalai.com/resources/glossaryhttps://safelegalai.com/resources/glossaryPlain-English definitions courts use when sanctioning AI-fabricated citations: hallucination, Hamid jurisdiction, wasted costs, Rule 11, and more.Thu, 16 Jul 2026 00:00:00 GMTexplainerglobalhttps://safelegalai.com/resources/readinghttps://safelegalai.com/resources/readingThe judgments, court rules, regulator guidance, and research worth reading first on AI in legal practice — annotated and grouped by who needs them.Thu, 16 Jul 2026 00:00:00 GMTroundupglobalhttps://safelegalai.com/analysis/pinsent-masons-sra-self-referralhttps://safelegalai.com/analysis/pinsent-masons-sra-self-referralIn Cork v Smith, a junior solicitor's AI tool fabricated the text of a rule that does not exist. Pinsent Masons self-referred to the SRA.Thu, 16 Jul 2026 00:00:00 GMTanalysisukhttps://safelegalai.com/analysis/pooja-ramesh-singh-india-supreme-court-aihttps://safelegalai.com/analysis/pooja-ramesh-singh-india-supreme-court-aiIndia's Supreme Court set aside two tribunal orders built on fake AI-generated precedents and ordered the Bar Council to frame AI rules for lawyers.Thu, 16 Jul 2026 00:00:00 GMTanalysisindiahttps://safelegalai.com/resources/guidance/singapore-ai-guidance-historyhttps://safelegalai.com/resources/guidance/singapore-ai-guidance-historySingapore's legal AI guidance, dated and sourced: the Supreme Court's 2024 circular, the Ministry of Law's 2026 sector guide, and the Law Society advisory.Thu, 16 Jul 2026 00:00:00 GMTguidance-diffsingaporehttps://safelegalai.com/analysis/south-africa-lpc-regulator-vacuumhttps://safelegalai.com/analysis/south-africa-lpc-regulator-vacuumSouth Africa's Legal Practice Council has published no guidance on lawyers' AI use — so its courts are referring practitioners to it instead, case by case.Thu, 16 Jul 2026 00:00:00 GMTanalysissouth-africahttps://safelegalai.com/analysis/three-years-legal-ai-hallucinations-2023-2026https://safelegalai.com/analysis/three-years-legal-ai-hallucinations-2023-2026From Mata v Avianca to a $110,000 Oregon sanction: what the verified record of AI-fabricated citations shows about courts, regulators, and the firms involved.Thu, 16 Jul 2026 00:00:00 GMTreportglobalhttps://safelegalai.com/resources/guidance/tier3-ai-guidance-india-brazil-hongkong-gulfhttps://safelegalai.com/resources/guidance/tier3-ai-guidance-india-brazil-hongkong-gulfAI guidance in India (Kerala), Brazil (CNJ Res. 615), Hong Kong, and the Gulf (QFC) — judicial and regulator documents beyond the big jurisdictions.Thu, 16 Jul 2026 00:00:00 GMTguidance-diffglobalhttps://safelegalai.com/resources/guidance/uk-judicial-ai-guidance-version-historyhttps://safelegalai.com/resources/guidance/uk-judicial-ai-guidance-version-historyHow the Courts and Tribunals Judiciary's AI guidance for judges in England and Wales changed across three versions: Dec 2023, April 2025, and 31 October 2025.Thu, 16 Jul 2026 00:00:00 GMTguidance-diffukhttps://safelegalai.com/resources/guidance/uk-legal-profession-ai-guidance-historyhttps://safelegalai.com/resources/guidance/uk-legal-profession-ai-guidance-historyThe version record of the SRA, Bar Council, and Law Society guidance on generative AI — what each says now, and when it changed.Thu, 16 Jul 2026 00:00:00 GMTguidance-diffukhttps://safelegalai.com/analysis/us-federal-court-ai-standing-ordershttps://safelegalai.com/analysis/us-federal-court-ai-standing-ordersNo national US rule governs AI in filings. A primary-sourced registry of representative federal judges' standing orders: certify, disclose, or prohibit.Thu, 16 Jul 2026 00:00:00 GMTreportus-federalhttps://safelegalai.com/analysis/mata-v-avianca-the-case-that-started-it-allhttps://safelegalai.com/analysis/mata-v-avianca-the-case-that-started-it-allTwo New York attorneys, six ChatGPT-invented cases, and a $5,000 sanction — the June 2023 ruling that created the legal AI hallucination era.Wed, 15 Jul 2026 00:00:00 GMTanalysisushttps://safelegalai.com/analysis/nsw-sc-gen-23-strictest-ai-court-rulehttps://safelegalai.com/analysis/nsw-sc-gen-23-strictest-ai-court-ruleNew South Wales runs the strictest AI court rule in the common-law world: affidavits must declare AI was NOT used. Here is what the practice note requires.Wed, 15 Jul 2026 00:00:00 GMTexplaineraustraliahttps://safelegalai.com/newsletter/this-week-in-legal-ai-safety-issue-0https://safelegalai.com/newsletter/this-week-in-legal-ai-safety-issue-0A sample issue showing the weekly format: incidents worldwide, regulator watch, one analysis take, and the tracker stat of the week.Wed, 15 Jul 2026 00:00:00 GMTnewsletterglobal