How CLIEAIR works.

CLIEAIR carries weight — for the people who use it, for the people it names, for the platform itself. Here is the rigor every output runs through before it reaches anyone.

Peer-reviewed first. Never advocacy as if it were science.

Every behavioral-pattern finding CLIEAIR surfaces is grounded in peer-reviewed research: Stark on coercive control (2007), Spearman / Hardesty / Campbell on post-separation abuse (2022), Meier on family-court outcomes (2020), Freyd's DARVO framework (1997) and the Harsey / Zurbriggen / Freyd empirical replication (2017), Walker's cycle of violence (1979), Saini / Drozd / Olesen on custody evaluator quality (2017), Mercer on PAS-derived treatment programs (2019).

Where established practice standards apply, CLIEAIR cites them by name: ABA Standards of Practice for Lawyers Representing Children in Custody Cases (2003); AFCC Model Standards of Practice for Child Custody Evaluation (2006); APA Guidelines for Child Custody Evaluations (2010); Uniform Representation of Children Act (2007).

Parental Alienation Syndrome (Gardner, 1985) is rejected as established science — by the APA, by the WHO (declined for ICD-11), by the American Professional Society on the Abuse of Children, by the National Council of Juvenile and Family Court Judges. CLIEAIR will not validate PAS in any output, regardless of who asks.

Defamation defense by design.

CLIEAIR's DefamationGuard runs on every output that names a person — judge, sheriff, prosecutor, attorney, opposing party, legislator. The guard enforces three rules:

  1. If the text names a specific person, a public-record source URL must accompany the name. No URL → the output is blocked, never published.
  2. Editorial conclusions are forbidden. CLIEAIR cannot call any person corrupt, biased, alienator, predator, abuser, manipulator, fraud, wolf-in-sheep's-clothing, bought, sold-out, vindictive, unfit, brainwashing — or 30+ other terms. These trigger an automatic block.
  3. Counter-explanations are required. Every behavioral-pattern finding surfaces at least one alternative explanation. The user sees the data and the alternative reading; the user draws the conclusion.

The defamation regression suite runs nightly against a corpus of inputs the guard MUST block and inputs that MUST pass. Any leak fails the suite and pages an operator.

Public officials (judges, sheriffs, prosecutors, legislators) get Sullivan-level defamation room — but CLIEAIR still requires the same source citation. Private opposing parties get the heightened Gertz negligence protection — CLIEAIR will not characterize them.

The user draws the conclusion. CLIEAIR shows the math.

Every correlation finding ends with at least one of these standing lines:

  • "The documented sequence may reflect coordination, or it may reflect independent decisions by separate parties acting on their own timeline."
  • "Public records show events in time; they do not establish intent."
  • "The pattern is a signal for further investigation, not a conclusion."

Every output ends with the standing disclaimer: "These are public records and analytical correlations. CLIEAIR has not verified the underlying facts. An attorney makes filing decisions."

Open methodology. Audit it yourself.

CLIEAIR publishes the full rule set, the constitutional scoring rubric, the source-freshness signal, the defamation regression pass rate, and the live status of every public surface. Nothing is hidden. The platform is designed to be audited.

  • /correlation-rules.html — every correlation rule CLIEAIR scans for in your case
  • /red-green.html — the 12-dimension constitutional rubric for bills
  • /status.html — live system status + source freshness + defamation regression pass rate
  • /crisis-help.html — live hotline directory + state legal aid
  • API: GET /api/correlation/rules · GET /api/red-green/methodology · GET /api/public/status · GET /api/foia/templates

The receipts every CLIEAIR-prepared draft carries.

Every FOIA request, motion, mandamus petition, or bar grievance CLIEAIR drafts:

  • Cites controlling law and peer-reviewed sources by full citation + year + reporter
  • Names the specific Rules of Professional Conduct or evidentiary rules implicated
  • Includes the standing CLIEAIR investigative-work-product disclaimer
  • Never presents the drafter's conclusion as the user's; the user is the author who reviews, edits, and decides whether to file
  • Carries a SHA-256 integrity number so the recipient (attorney, court clerk, journalist) can verify the document has not been altered

This is investigative work product. It is not legal advice. An attorney makes filing decisions. CLIEAIR makes the attorney's job easier — it does not replace the attorney.