AI Policy

ReaDDy uses Google Gemini, a large language model (LLM) developed by Google, to generate critique and scores for your essay answers. This document explains how AI is used, what its limitations are, and what you can and cannot rely on it for.

When you request a critique of a submitted essay answer, our AI system:

• Evaluates your answer against the ALAC method (Answer, Legal Basis, Application, Conclusion) — the writing framework prescribed by the ICAN Bar Coaching Program and aligned with Philippine Bar Examination expectations.
• Scores your answer on the ABCS rubric (Accuracy, Brevity, Clarity, Suavity, Time Management) and the Case Problem Analysis rubric (Facts, Legal Basis, Conclusion), each on a 1–5 scale.
• Checks whether expected keywords, doctrines, and case citations are present in your answer.
• Provides an overall score out of 5, constructive feedback, identified gaps, and a coaching comment.
• Compares your answer against reference suggested answers (which are never shown to you directly).

AI-generated critique on ReaDDy is a training aid only. It is not:

• Not a prediction of your actual bar exam score or result.
• Not legal advice. Nothing in AI feedback constitutes legal opinion or professional advice from a lawyer.
• Not infallible. AI may make errors in identifying correct legal principles, miss nuance in complex doctrines, or misread the specific legal issue in a question.
• Not a substitute for your review center, professor, or licensed bar reviewer.
• Not authoritative on Philippine jurisprudence — always verify doctrines and case holdings with primary legal sources.

When you request a critique, the following data is sent to Google Gemini: the bar subject, the question text, your essay answer, the keywords/doctrines associated with the question, and suggested reference answers (server-side only). Your name, email, and account details are never sent to the AI. Only the content needed to evaluate the specific answer is transmitted. Google processes this data under its own AI usage policies and does not use it to train Gemini models without opt-in consent (per Google's standard API terms).

To prevent abuse, AI critiques are limited per account tier per day (5 critiques/day on free trial; 20 critiques/day on Bar Season Pass and center-enrolled accounts). These limits reset daily at midnight PHT (Asia/Manila). If you reach your daily limit, you will see a friendly prompt to upgrade or wait until the next day.

Students enrolled under a review center may also receive human feedback from their assigned reviewer. Reviewer annotations, scores, and remarks are independent of and in addition to AI critique. For assigned exams, AI critique is disabled entirely — only reviewer feedback applies. Where human feedback and AI feedback conflict, the human reviewer's judgment takes precedence.

We continuously improve the AI critique system by refining prompts, updating the rubric, and expanding the keyword database for each bar subject. However, we make no warranty that AI critique is accurate, complete, or up-to-date with the latest Supreme Court decisions. If you believe an AI critique contains a significant error in legal analysis, please report it to readdy@ddlaw.ph and we will review it.

By creating a ReaDDy account, you consent to your essay answers being processed by our AI system for the purpose of generating critique and scores as described in this policy. You may choose not to request AI critique on any individual answer — the "Critique my answer" button is opt-in per question. AI critique results are stored in your account alongside your submitted answer.

We may update this AI Policy as the technology or our practices change. We will notify registered users by email or in-app notice when material changes are made.

Questions about our use of AI: readdy@ddlaw.ph