Rahmat Wibowo posted an AI 'analysis' rating Aqilah Saffanah's chat messages at 'IQ 63', 'mental age 15.8' and 'primary school' education, linking the full report on his company's GitHub, while warning others not to 'weaponize pseudo-AI analysis'

Original post ↗

Rahmat Wibowo posted an AI 'analysis' rating Aqilah Saffanah's chat messages at 'IQ 63', 'mental age 15.8' and 'primary school' education, linking the full report on his company's GitHub, while warning others not to 'weaponize pseudo-AI analysis'

Commentary

He posted it on LinkedIn on 18 May 2026.

  • Whose report? He presents the numbers as a report he “came across”. In the comments he links the “full report” himself, in his company InfraLoka’s GitHub repository for an “IQ agent”, in a folder named “aqilah”, with the file name “2026-05-18-aqilah-analysis-report”.
  • “IQ 63”, “mental age 15.8”, “primary school”. He posts the scores against Aqilah Saffanah’s name while quoting the report’s own warning that attributing them “to any specific person… is methodologically invalid”, and that broken tokenization and poor Indonesian support “heavily skewed the results”.
  • “Don’t weaponize pseudo-AI analysis.” The post’s lesson is that people who “weaponize pseudo-AI ‘analysis’ to attack others’ intelligence” reveal more about themselves, and that one should respect boundaries “instead of doxxing employers”. In the comments he tags her employer ParagonCorp and its CEO Salman Subakat, a law firm and the bar association PERADI.
  • The start of a pattern. Four weeks later he accused her, without evidence, of running an anonymous account against him. In November he posted a private chat with her and accused her of insulting Islam, again tagging ParagonCorp.

Transcript

[LinkedIn post — Rahmat Wibowo] Important Reflection on Digital Harassment & Misuse of "AI Analysis" Tools Today, I came across a detailed text analysis report on a WhatsApp/group conversation thread involving coordinated harassment, legal threats, doxxing attempts, and reputational attacks against a targeted individual. Key takeaways from the professional analysis (dated May 18, 2026): Word count 2,837 VFS 16.3 / 100 IQ estimate 63 "Low" (CI 57–69) Mental age 15.8 (@ assumed age 25) Education SD / Primary School (MEDIUM confidence) Major tool artifacts heavily skewed the results: broken tokenization of timestamps, emojis, and metadata led to near-zero lexical scores. Indonesian language support in the parser was also insufficient. The report repeatedly warns: "Attribution of VFS, IQ, or education estimates to any specific person... is methodologically invalid." Despite these clear disclaimers, certain individuals (including someone named Aqilah Saffanah) appear to have been part of a group effort pushing mass reporting, media pressure, and psychiatric stigmatization ("gila", "cacat mental", etc.) across platforms. This highlights a broader issue: When people weaponize pseudo-AI "analysis" to attack others' intelligence or character — especially while engaging in coordinated harassment themselves — it reveals far more about the attackers than the target. […] Respecting boundaries instead of doxxing employers, family contacts, or personal addresses […] #DigitalEthics #OnlineHarassment #ResponsibleAI #MentalHealthAwareness #Indonesia [Comment — Rahmat Wibowo] full report https://github.com/InfraLoka/verbal-fluency-iq-agent-skills/blob/main/example/aqilah/report/2026-05-18-aqilah-analysis-report.md [Comment — Rahmat Wibowo] ParagonCorp Salman Subakat Salman Subakat [Comment — Rahmat Wibowo] Ahmad Fikri Assegaf Assegaf Hamzah & Partners Assegaf Kawilarang & Associates Assegaf & Justicio Law Firm [Comment — Rahmat Wibowo] PERADI DPN PERADI Perhimpunan Advokat Indonesia (PERADI)