Reliable Bias Articulator

Authorship · Integrity · Environment

A commitment before you begin

The words we write now
carry a planetary weight.

Every AI query draws from a watershed and a power grid. This platform measures authorship, surfaces environmental cost, and helps institutions understand the full weight of AI-assisted creation, including how confident the analysis itself can be.

~0.012gallons of water
per AI query

~0.003watt-hours per
completion

100Ksgallons to train
one LLM

I acknowledge that AI-assisted writing consumes water and energy from natural systems, scaling with query volume. I understand that data centers draw from regional watersheds and power grids with real community impacts. I commit to using this tool with intentionality , to illuminate my authorship, not obscure it. I recognize that authentic human authorship has intrinsic value to literary culture and academic integrity.

Reliable Bias Articulator

Authorship transparencyinfrastructure.

For writers, students, educators, and institutions navigating the disclosure landscape, built without judgment and designed for honesty.

Nothing you submit is stored. Work samples enter the analysis and return to you.

Disclosure is a professional act, not a confession. This tool is designed accordingly.

The confidence rating is as important as the score.

Tier adjustments can be made month to month from your client portal, with 48 hours notice before your next billing cycle.

Choose your access tier

Trial 30 days free

$0 for 30 days

Full access to explore every feature before choosing a plan. No credit card required.

All three analysis modes
Process map and disclosure framing
Copyright similarity analysis
Community Connector
100 total analyses
No card required

Free

$0 / month

Essential access for occasional use.

Copyright similarity analysis
Community Connector
Privacy and data notice
5 authorship analyses / month
No process map export

Individual

$14 / month

For working writers, students, and independent practitioners.

All three analysis modes
Process map and disclosure framing
25 analyses / month included
$0.25 per additional analysis
$3.00 per additional user seat

Institution Institutional

$150 / month

For literary journals, writing programmes, contest organisations, and academic departments.

All three analysis modes
50 user seats included
100 analyses / month
$0.25 per additional analysis
$10 per additional user seat
Process map and disclosure framing
Community Connector
Priority support

Tier adjustments can be made month to month from your client portal. Changes must be submitted at least 48 hours before your next billing cycle to take effect that month. Overage charges ($0.25 per additional analysis) are billed at the end of each cycle.

Account details

First name

Last name

Email address

City

State

Phone number (optional)

Set a password

Password

Confirm password

Your work samples are never stored. What you submit for analysis passes through Anthropic's API and returns to your browser. Account details (name, email, city, state, phone) are stored securely to manage your tier and billing. See the Privacy and Data Notice.

By creating an account you agree to the RBA Terms of Use. Paid tier billing is handled securely via Stripe.

Email address

Password Forgot password?

No account? Create one.

See how it works

Four use cases, select one to preview

Standard mode

Eight-phase process record

Set AI involvement per phase with sliders. Output: scored result, process map, disclosure framing.

Streamlined mode

Describe, don't configure

Write your process in plain language. The engine infers phase scores. Confidence reflects your specificity.

Stylistic similarity analysis

Genre-calibrated. Each signal rated for confidence. Developmental recommendations, not legal verdicts.

Community connector

Find your literary community

ZIP-code search for nonpartisan literary and conservation organisations near you.

Authorship integrity platform · Academic & literary · Est. 2026

Who wrote this?Every word. Confidence rated.

Three modes of authorship analysis. A copyright similarity check. An interactive process map. And now: a confidence rating on every result, from highly confident to conflicted.

Choose analysis mode Copyright check

3Analysis modes

3Confidence levels

8Phases mapped

H₂OWater footprint shown

Choose your analysis mode

Three ways to articulate authorship

Each mode returns a confidence rating on a spectrum from Highly Conflicted Analysis (pale pink) through Moderately Confident Analysis (yellow) to Highly Confident Analysis (green), surfacing the limits of the analysis alongside its findings.

Standard

Detailed Phase Input

Full phase-by-phase slider control, tool declaration, and submission context. Best for course submissions and institutional audits.

Streamlined

Automated Analysis

Five inputs only. AI infers phase involvement from your process description and sample. Best for journal submissions and grant applications.

Stylistic and linguistic similarity analysis against known copyrighted work traditions, with confidence-rated originality findings.

Standard mode

Detailed Phase Control

You set AI involvement per phase. The engine applies genre-weighting and returns a confidence-rated analysis.

Phase-by-phase sliders

Set AI involvement across all 8 phases from 0 to 10.

Tool declaration

Specify which AI tools were used for institutional record.

Confidence rating

Green / yellow / pale pink signal how certain the analysis is, with a detailed breakdown of contributing factors.

Process map upgrade

Color-coded flow of human vs. AI interaction across all phases.

Process Map Upgrade

Interactive color-coded map of human–AI interaction across all 8 phases.

Enable

Author name

Genre

Context (optional)

Approximate word count Optional · rough estimate is fine (±1,000 for long-form · ±30 for poetry)

Work sample Published or unpublished, any length

Process notes (optional)

AI involvement by phase (0=none · 10=fully AI)

Ideation / planning

Research

Rough draft

Revision

Editing / proofread

Line / structural

Query / pitch

Marketing

AI tools used

Analyzing…

Streamlined mode

Five inputs. Full intelligence.

Phase scores inferred automatically. Confidence rating reflects how much signal the process description provides.

Inferred phase scoring

AI reads your process description and assigns phase scores.

Voice analysis

Sample assessed for stylistic consistency and authentic voice markers.

Confidence rating

Shorter or vaguer process descriptions typically yield yellow or pale pink confidence. Longer, specific descriptions earn green.

Process map upgrade

Inferred phases visualized as a color-coded human–AI flow map.

Process Map Upgrade

Visualize inferred human–AI interaction as a color-coded flow map.

Enable

Author name

Genre

Approximate word count Optional · rough estimate is fine (±1,000 for long-form · ±30 for poetry)

Work sample Published or unpublished, any length

Process description (max 300 words)

0 / 300 words

Analyzing…

Copyright Similarity Check

Full copyright analysis with confidence rating is in the dedicated section below.

Go to Copyright Check ↓

How original is your work?

Stylistic, structural, and linguistic similarity analysis, with a confidence rating on every finding. A literary analysis aid, not a legal determination.

Stylistic similarity

Sentence structure, lexical choices, and narrative rhythm echoing known works.

Genre-specific signals

Risk calibrated for your genre; poetry risks differ from long-form fiction.

Confidence-rated findings

Each similarity signal rated green / yellow / pale pink by the engine's certainty.

Originality recommendations

Four targeted suggestions, developmental and non-punitive.

Disclaimer

This is a literary analysis aid, not legal advice. Always consult qualified legal counsel for copyright matters.

Work details

Author name

Genre

Known influences (optional)

Work sample Published or unpublished · 80–600 words recommended

Context notes (optional)

Analyzing for similarity…

The cost beneath the cursor

Writing now uses water.

H₂O~0.012Gallons per queryData center cooling water per LLM completion

kWh~0.003Watt-hours per callFrom regional grids, often fossil-fuel sourced

CO₂100KsGallons per LLM trainedThousands of swimming pools per model

Community Connector

Find nonpartisan literary and conservation organizations near you, searchable by zip code.

Open directory →

Privacy & Data Notice

What this tool does and does not do with your work samples. No ambiguity, no boilerplate.

Read the notice →