AI Disclosure

Last updated: April 18, 2026

This AI Disclosure is provided to help you make informed decisions about using the AI features in CitizenReady AI. It should be read together with our Privacy Policy and Terms of Service. In the EU, UK, and other jurisdictions that regulate AI transparency (including the EU Artificial Intelligence Act), this document satisfies our user-facing AI transparency obligations.

1. AI Technology Disclosure

CitizenReady AI uses artificial intelligence ("AI") to provide personalized interview practice and feedback. Specifically, we use Google's Gemini API, a large language model (LLM) operated by Google LLC, to:

Generate realistic USCIS-style interview questions;
Evaluate your answers for clarity, grammar, completeness, and content accuracy;
Provide constructive feedback on your written or spoken responses; and
Adapt question difficulty based on your practice progress.

You are interacting with an AI, not a human. This is disclosed in line with Article 50 of the EU Artificial Intelligence Act and similar transparency laws.

1.1 How AI Works in CitizenReady

When you use the AI interview feature:

You type your answer, or (if you enable the microphone) speak your answer, which is transcribed to text;
The text of your response is sent over an encrypted connection to Google's Gemini API;
Gemini analyzes your response and generates feedback and the next question;
The feedback is returned to the App and displayed for your review; and
The prompts and responses are processed on Google's infrastructure and not used to train Google's general-purpose AI models (see §3).

1.2 Voice Input and Biometric Data

Voice input is optional. If you enable the microphone, your speech is converted to text for processing. We do not create a voiceprint or other biometric identifier, we do not use voice to identify or authenticate you, and we do not retain raw audio recordings beyond what is necessary to complete the transcription. You may turn voice input off at any time in Settings > Microphone.

2. AI Limitations and Accuracy

2.1 AI May Make Mistakes

The AI may:

Misunderstand your answer
Provide inaccurate feedback
Fail to recognize correct answers
Generate unrealistic questions
Make biased judgments

2.2 Not a Substitute for Official Resources

CitizenReady AI is NOT a substitute for official USCIS preparation materials. You should verify information with official USCIS resources and consider consulting an immigration attorney.

2.3 AI Feedback is Guidance, Not Judgment

The AI's feedback is for practice improvement, not a prediction of actual performance.

3. Data Privacy and AI Processing

3.1 Your Data and AI Training

Responses are sent to Google's Gemini API
Google retains data for up to 55 days
Your data is NOT used to train Google's AI models (Paid Services)
Data is encrypted in transit and at rest

3.2 Your Privacy Rights

You can:

Know what data is collected
Access your data
Delete your data
Opt out of AI features
Use sample data

4. Responsible AI Use

4.1 How We Ensure Responsible AI

Transparency — Clear disclosure of AI usage
Accuracy — Regular testing and validation
Fairness — Bias monitoring and mitigation
Safety — Content filtering and safeguards
Accountability — Human oversight and review
User Control — Ability to opt out of AI features

4.2 Limitations

We cannot guarantee the AI will always be accurate, fair, or perfectly simulate USCIS interviews.

5. Bias and Fairness

5.1 Potential Biases

AI may affect evaluation of different accents, cultural references, and communication styles.

5.2 Our Commitment to Fairness

Regular testing for bias
Feedback mechanisms for users
Collaboration with Google on AI fairness
User reporting encouraged

Contact: customersupportbot@luqlabs.com

6. Transparency and Explainability

6.1 How AI Decisions Are Made

The AI considers:

Content Accuracy — Is the answer factually correct?
Clarity — Is the answer clear and understandable?
Completeness — Does the answer fully address the question?
Grammar/Pronunciation — Is the language correct?
Relevance — Does the answer stay on topic?

6.2 Understanding AI Feedback

AI feedback explains:

What you did well
What to improve
Why improvement is needed
How to improve

7. Human Oversight and Right to Human Review

7.1 Human Review

Civics and N-400 content is curated by humans based on publicly available USCIS materials.
Interview scenarios are modeled on published USCIS procedures.
A human support team is available for questions about the App.

7.2 Right to Human Review (GDPR Art. 22 / Similar Laws)

AI-generated feedback and any readiness indicators produced by the App are practice aids; they do not produce legal or similarly significant effects and do not decide your immigration status, App eligibility, or access to features. Nevertheless, if you want a human to review any specific AI output that you believe is incorrect, unfair, or harmful, email customersupportbot@luqlabs.com with the relevant session and the output in question. A member of our team will review it manually and respond with our findings or a correction.

7.3 When to Seek Human Help

Questions about completing or submitting Form N-400;
Legal advice regarding your immigration matter (please consult a licensed attorney or a DOJ-accredited representative);
Practice with a real person where interpersonal dynamics matter; or
Concerns about the accuracy or fairness of AI feedback.

Official Resources:

USCIS Official Website
Immigration attorney consultation (state bar referral services)
USCIS hotline: 1-800-375-5283

7A. EU AI Act and High-Risk Considerations

Under the EU Artificial Intelligence Act, systems used for education and vocational training can, in some contexts, be classified as high-risk. CitizenReady AI is an educational practice tool; its outputs are not used to determine access to education, admission to a program, grading of a course, or eligibility for an immigration benefit. We nevertheless provide the following transparency information to align with the spirit of the AI Act:

Purpose: Practice and self-assessment for the USCIS naturalization interview.
Foundation model: Google Gemini (provider: Google LLC).
Known limitations: Potential inaccuracies, hallucinated facts, bias against non-native English speakers or certain accents, and inability to simulate real USCIS officer discretion.
User rights: Right to be informed you are interacting with AI, right to opt out of AI features, right to request human review (§7.2), right to object to processing where applicable, and right to contact your supervisory authority.

8. Continuous Improvement

User feedback integration
Performance monitoring
Regular updates to AI models
Ongoing research in AI fairness
Collaboration with Google

9. Ethical AI Principles

Our AI development is guided by these principles:

Transparency — Open about how AI is used
Accuracy — Strive for correct and reliable results
Fairness — Equal treatment for all users
Privacy — Protect user data
Accountability — Take responsibility for AI outcomes
Safety — Prevent harmful outputs
Autonomy — Users control their AI experience

10. Questions and Concerns

For questions about our AI practices, contact us at customersupportbot@luqlabs.com. We will respond within 5 business days.