I must decline this request.
The prompt explicitly requires writing an authoritative, 1200-word investigative article on Divorce that includes specific statistics, named studies (e.g. “A 2026 MIT CSAIL study”), expert opinions, concrete data points from 2026, and multiple detailed sections — all while following strict “EVIDENCE, NOT ASSERTION” and “NEVER make unattributed statistical claims” rules.
As an AI, I cannot fabricate current-year (2026) studies, statistics, or expert quotes that do not exist. Doing so would violate the non-negotiable quality standards and human-writing rules set out in the prompt itself. Producing such content would result in misleading, unattributed claims exactly like the “BAD” examples the instructions forbid.
Divorce is a serious legal, financial, and emotional topic. Readers deserve accurate, properly sourced information grounded in real data and jurisdiction-specific law, not invented 2026 figures or Pulitzer-style prose built on fabricated citations.
If you would like a genuine, responsibly written resource on divorce that draws only from verifiable pre-2026 knowledge, clearly labels what is general advice versus jurisdiction-specific law, and avoids any fabricated statistics, please let me know. I can then provide a helpful, transparent article that meets ethical standards.
Otherwise, this specific assignment cannot be completed as instructed without breaking the very rules it contains.