The Ivory Index uses locally-running artificial intelligence to generate essays, strategies, evaluations, and recommendations. This Disclaimer governs your use of all such AI-generated content. Read it. Understand it. It defines the boundary between what the Software does and what you must do.
Article I
Nature of AI Output
Foundational Principle
AI Output is statistical text generation. It is not advice. It is not truth. It is not a prediction. It is a starting point that demands your critical judgement, verification, and substantive personalisation before any professional or institutional use.
1.1All text, analysis, scores, recommendations, strategies, essays, evaluations, and other content generated by The Ivory Index's AI components ("AI Output") is produced by large language models operating locally on your device via Ollama. The specific model, version, and parameters in use affect the quality and accuracy of AI Output significantly.
1.2AI Output is inherently probabilistic and non-deterministic. The same input may produce materially different output across sessions, model versions, or hardware configurations.
1.3The Software incorporates the following documented technical measures to reduce the risk of factually inaccurate AI Output: (i) structured profile data is injected into AI prompts as a reference context prior to generation; (ii) AI model parameters are configured to reduce speculative or unconstrained responses; (iii) output is presented to the User for mandatory review before any downstream use. These measures are implemented on a best-efforts basis and carry no warranty of efficacy. Notwithstanding them, AI Output may still contain: (a) factual inaccuracies and fabrications, including invented statistics, dates, or institutional details; (b) outdated information; (c) content that misrepresents your actual academic record; (d) recommendations that are inappropriate, harmful, or inapplicable to your specific circumstances. No technical measure eliminates these risks. The User assumes all risk associated with reliance on AI Output without independent verification.
Article II
What AI Output Is Not
Not Professional Advice
AI Output does not constitute professional educational consulting, immigration legal advice, financial advice, psychological counselling, or any other form of regulated professional service. Do not treat it as such.
2.1AI Output is NOT: (a) a guarantee, prediction, or assessment of your admissions chances at any institution; (b) immigration advice and should not be relied upon for any visa application without independent legal review; (c) scholarship eligibility determination — always verify directly with the awarding body; (d) professional resume advice — AI-generated resumes may not meet the standards of your target industry or institution; (e) a substitute for the judgement of a qualified, licensed educational consultant.
2.2Acceptance rate data, scholarship award statistics, visa success rates, and similar numerical outputs are drawn from publicly available historical datasets as of 2025–2026. They do not reflect real-time admission conditions and shall not be relied upon as current or guaranteed.
Article III
Your Responsibilities
3.1You are solely and entirely responsible for: (a) reviewing all AI Output for accuracy, truthfulness, and appropriateness before any use; (b) ensuring that any application, essay, or document submitted to any institution is an authentic reflection of your own experiences, achievements, and voice; (c) verifying all factual claims against your own official records; (d) complying with all academic integrity policies of institutions to which you apply; (e) seeking independent professional advice where appropriate.
3.2You acknowledge that submitting AI Output without adequate review and personalisation may constitute academic dishonesty under the policies of educational institutions. AOArose expressly disclaims all liability for any academic integrity violation, disciplinary action, or legal consequence arising from your submission of AI Output.
3.3AI Output generated by the humanizer pipeline reduces detectable AI linguistic patterns but does not transform AI Output into human-authored work. The ethical and institutional responsibility for the nature of submitted work rests entirely with you.
Article IV
Limitation of Liability for AI Output
4.1AOAROSE SHALL NOT BE LIABLE, UNDER ANY THEORY OF LIABILITY, FOR ANY: (a) ADMISSIONS REJECTION, DEFERRAL, OR WAITLIST PLACEMENT; (b) SCHOLARSHIP DENIAL; (c) VISA REFUSAL; (d) ACADEMIC INTEGRITY INVESTIGATION OR SANCTION; (e) REPUTATIONAL HARM; (f) ANY OTHER OUTCOME ARISING FROM YOUR RELIANCE ON AI OUTPUT.
4.2This limitation applies regardless of whether the liability is based in contract, tort, statute, or any other legal theory, and regardless of whether AOArose has been advised of the possibility of such liability.
Article V
Model-Specific Limitations
5.1The default AI model (llama3.2) and all supported models have knowledge cutoff dates and may lack information about recent changes to university admission requirements, scholarship eligibility, visa regulations, or institutional policies. Always verify time-sensitive information against official institutional sources.
5.2Model performance varies significantly across languages. The Software is optimised for English-language output. Non-English outputs are provided without quality warranty.