Rahmat Wibowo posted an article about AI-generated legal responses containing critical flaws, then used it as a platform to call for a review of a specific individual (Albert Suwandhi) while naming other individuals in what appears to be a defamatory manner.

| just published a research article on something most people don't talk about. What happens when someone uses Al to respond to a formal legal notice — without a real lawyer? I witnessed it firsthand in a dispute I'm personally involved in. The response looked impressive: constitutional citations, Latin legal terminology, peer-reviewed NIH references. Three languages of authority. It was, on forensic examination, a near- complete gift to the opposing party. Five failure modes | documented: 01. Admitting mens rea while claiming free speech protection — calling posts "sarcastic criticism” inadvertently confirmed the deliberate intent required under Art. 27A UU ITE 02. Citing Art. 28E (free expression) while ignoring Art. 28G (right to personal honor) — half a constitutional argument is worse than none 03. Deploying accurate science to rebut a claim that was never made — three bipolar heritability studies, zero relevance to the actual legal standard 04. Getting burden of proof half right — actori incumbit probatio is criminal doctrine; the civil standard under Art. 1365 KUH Perdata is different 05. The chronological own-goal — confirming the posts were never deleted while arguing there was no ongoing harm The pattern: Al produces text that is locally plausible but globally incoherent. Rhetorically confident but legally catastrophic. | build Al infrastructure for a living. | believe in what it can do. But there is a hard line between research assistance and legal strategy — and crossing it in formal proceedings is permanently costly. ‘Tagging two people whose work sits at the center of this: Pramudya A. Oktavinanda— the governance gap here is real. Al tools generating plausible legal text without jurisdictional guardrails are a systemic risk in Indonesia's UU ITE + UU PDP landscape. Abhishek Gupta — this case is a textbook illustration of Al overconfidence without human oversight. The Indonesian hybrid civil law context makes the failure modes even more pronounced. Brian Finlay — this case is a textbook illustration of Al overconfidence without human oversight. The Indonesian hybrid civil law context makes the failure modes even more pronounced. The Stimson Center Five failure modes in detail, a framework for responsible Al-assisted legal work, and practical recommendations for individuals and organizations. The algorithm played lawyer. It should not have. Dedicated to Steven Nataniel Kodyat from Xendit Moses Lo Miki Burman N. Willy PT #ResponsibleAl #AlGovernance #LegalTech #AIRisk #IndonesiaLaw #UUITE #AlEthics #infraloka #AlAccountability