Our Methodology & Transparency

1. What "Forensic" Actually Means

RawReach uses the word "forensic" as a metaphorical marketing term. It describes our process of close, detail-oriented examination of written communication — much the same way "surgical" precision or "deep-dive" analysis are used in marketing to convey thoroughness.

Important: RawReach is not affiliated with, and does not perform, forensic science, forensic psychology, forensic linguistics (as a credentialed discipline), criminal investigation, psychological profiling, or any form of legal or clinical assessment. We are a marketing SaaS platform.

Wherever the word "forensic" appears in our platform, copy, or outputs, it is always paired with its accurate technical description:

"Forensic" = Behavioral Pattern Analysis — automated linguistic pattern recognition applied to written text samples for the purpose of generating marketing copy. No licensed professional is involved. No clinical judgment is rendered.

If you arrived at RawReach expecting forensic psychology services, legal evidence analysis, or clinical personality assessment, please be advised that RawReach does not offer those services. Our outputs are marketing content, not professional reports.

FTC March 2026 AB 489

2. About Our AI Logic (AB 853 / ADMT 2026)

Pursuant to California AB 853 and the Automated Decision-Making Technology (ADMT) regulations effective 2026, RawReach discloses the following about its AI processing pipeline.

Our AI system is named RawReach AI — Behavioral Pattern Logic v3.1. Every piece of content generated through the platform is labeled with this identifier so you always know exactly what produced it.

⬡

AI Transparency Label (appears on every output) "Analyzed by RawReach AI — Behavioral Pattern Logic v3.1 | For marketing use only. Not a clinical diagnosis or legal assessment."

The ADMT pipeline operates through three sequential gates:

GATE 1
De-Identification All personal identifiers — names, phone numbers, email addresses, account handles, and any direct identifiers — are stripped before text enters the analysis pipeline. Processing occurs on anonymized data only. No personal data is retained after this gate clears.
GATE 2
Behavioral Domain Classification The de-identified text is analyzed for: conflict typology (the category of interpersonal dynamic present in the writing), communication patterns (passive, direct, avoidant, aggressive, etc.), motivational drivers (validation-seeking, autonomy, reciprocity, etc.), and narrative framing (victim, protagonist, mediator, etc.). These are probabilistic classifications, not definitive assessments.
GATE 3
Content Generation Based on the domain classifications from Gate 2, the system generates marketing copy tailored to the conflict niche. All output is clearly labeled as AI-generated marketing content. No output should be construed as advice, diagnosis, or professional opinion of any kind.

You have the right to request a plain-language explanation of any decision made by this system that affects you. See Section 5 for how to exercise your rights.

AB 853 ADMT 2026

3. What We Analyze

RawReach's Behavioral Pattern Analysis (our "forensic" process) examines linguistic patterns in text only. The following list is exhaustive — we analyze nothing beyond it.

What we analyze:

Communication style — formality, directness, emotional tone in writing
Conflict resolution patterns — how disputes are framed and addressed in text
Value signals — recurring themes, priorities, and concerns expressed in writing
Narrative framing — the role the writer assigns themselves and others in their account

What we do not analyze, determine, or provide:

Psychological diagnosis or assessment of any kind
Legal determinations, evidence analysis, or forensic opinions
Clinical assessment (personality disorders, mental health conditions, etc.)
Predictions of future behavior
Credibility assessments ("is this person lying")
Medical, therapeutic, or counseling guidance

All RawReach outputs are probabilistic marketing content patterns. They represent statistically common language approaches associated with the detected communication style. They are not facts about any individual, and should not be treated as such.

FTC March 2026 AB 489 ADMT 2026

4. Zero-Footprint & Split-Storage Architecture

RawReach is built on a zero-footprint principle: we collect only what is necessary to deliver a result, and we destroy raw inputs immediately upon delivery.

Zero-Retention: All raw text inputs are permanently and irreversibly purged upon content delivery. We do not store, archive, or sell your input data.
Split-Storage: Anonymized behavioral classification tags (Gate 2 output) are stored separately from any account or billing data, with no linkable bridge between them.
No Re-Identification: We do not attempt to re-associate de-identified analysis outputs with any individual.
California Zero-Retention Compliance: Our architecture is designed to satisfy California's Zero-Retention standard for sensitive personal information processed by AI systems.

For complete data handling details, see our Privacy Policy and Terms of Service.

5. Your Rights Under ADMT 2026

California's Automated Decision-Making Technology regulations (effective 2026) grant you the following rights with respect to any AI-driven processing RawReach performs.

Right to Access Logic You may request a plain-language explanation of the automated logic that produced any specific output or classification. Contact us at hello@rawreach.co with subject line "ADMT Logic Request." We will respond within 10 business days.

Right to Opt Out You may opt out of having your text processed by automated decision-making systems at any time. If you opt out, RawReach cannot generate content for you, as ADMT is core to our service. To opt out, email hello@rawreach.co with subject line "ADMT Opt Out."

Right to Human Review You may request that a human review any automated output that you believe is inaccurate or has adversely affected you. We will arrange a human review within 15 business days of your request. Note: human reviewers can assess whether the AI followed its documented logic; they cannot retroactively access your raw input text, which has been purged.

To exercise any of these rights, email hello@rawreach.co or write to: RawReach, Silicon Valley, CA.

ADMT 2026 AB 853

6. FTC & AB 489 Compliance

RawReach complies with FTC guidance (March 2026) on AI-generated content and California AB 489 governing marketing claims about AI systems.

Hedging Language Policy: All RawReach marketing copy and platform outputs follow a mandatory hedging policy. No output makes objective, definitive, or clinical claims about any individual or group. All outputs use probabilistic language including but not limited to: "may suggest," "patterns associated with," "commonly indicates," "tends to correlate with."

No Objective Claims: RawReach does not claim that its Behavioral Pattern Analysis produces objective truth. Our outputs represent patterns statistically associated with detected linguistic features — not factual determinations about any person's character, psychology, or behavior.

Marketing Use Only: All outputs are designated for marketing copy generation only. RawReach expressly prohibits use of its outputs as evidence in legal proceedings, as basis for medical or psychological decisions, as justification for employment decisions, or as support for any decision that materially affects a person's rights or opportunities.

Disclosure required by AB 489: This platform uses automated AI systems to analyze text and generate marketing content. The AI system (RawReach AI — Behavioral Pattern Logic v3.1) has not been validated for clinical, legal, or forensic professional use. Outputs are probabilistic patterns, not professional opinions.

FTC March 2026 AB 489 AB 853 ADMT 2026