🤖

AI Transparency & Responsible Use

How we use artificial intelligence ethically and responsibly

1. AI Systems We Use

QuickPetition uses the following AI systems to provide our immigration petition drafting service:

🧠 Primary AI Language Model

Purpose: Generate petition letter drafts, analyze qualifications, provide recommendations
Type: Advanced large language model
Provider: Third-party enterprise AI service
Data Processing: Your information is sent to the provider's API for processing
Data Retention: Provider does not retain data sent via API (as per enterprise terms)

💬 Secondary AI Language Model

Purpose: Alternative petition generation, formatting, content refinement
Type: Advanced large language model
Provider: Third-party enterprise AI service
Data Processing: Your information is sent to the provider's API for processing
Data Retention: Provider does not train on API data (zero data retention policy)

2. EU AI Act Compliance

The European Union's AI Act (Regulation 2024/1689) classifies AI systems by risk level. QuickPetition's compliance status:

⚠️ Risk Classification: Limited Risk

Our AI system is classified as Limited Risk under the EU AI Act, requiring transparency obligations but not full high-risk compliance.

Why Limited Risk? We provide AI-generated content (petition drafts) that users review before submission. We do not make automated decisions about immigration outcomes, biometric identification, or critical infrastructure.

Our EU AI Act Compliance Measures:

Transparency: Clear disclosure that content is AI-generated
Human Oversight: Users must review all AI output before use
No Automated Decisions: Final immigration decisions made by USCIS, not our AI
Data Quality: Users responsible for providing accurate input data
Risk Management: Regular audits of AI output quality and bias
Record Keeping: Logs of AI generations for accountability (90 days)

3. Algorithmic Transparency (GDPR Article 22)

Under GDPR Article 22, you have the right not to be subject to fully automated decision-making. QuickPetition complies:

✅ No Fully Automated Decisions

Our AI generates suggestions and drafts, not final decisions
You retain full control to review, edit, and approve all content
Immigration outcomes are determined by USCIS officers, not our AI
You can opt out of AI-generated content and use manual drafting

How Our AI Works:

Input Collection: You provide information about qualifications, achievements, publications
AI Processing: Third-party AI language models analyze your profile against EB-1A/EB-NIW criteria
Draft Generation: AI creates a petition letter draft highlighting your strengths
Human Review: You review, verify accuracy, and make edits
Final Approval: You decide whether to use the draft (no automated approval)

4. AI Ethics & Bias Mitigation

Our Ethical AI Principles:

Fairness: AI models trained on diverse immigration cases, not biased toward specific nationalities
Accuracy: Regular quality checks to ensure AI output aligns with USCIS requirements
Transparency: Clear labeling of AI-generated content (never hidden)
Accountability: Human oversight for all AI outputs before delivery to users
Privacy: No AI training on your personal data without explicit consent
Explainability: AI provides reasoning for recommendations

Bias Prevention Measures:

Regular audits for nationality, gender, or ethnic bias in AI outputs
Diverse prompt engineering to avoid stereotyping
User feedback mechanisms to report biased or unfair outputs
Third-party AI providers have their own bias mitigation protocols
No use of protected characteristics (race, religion) in AI decision logic

5. Data Usage for AI Training

🔒 Your Data is NOT Used for Training

We do NOT use your personal information to train AI models. Your petition data, documents, and personal details are:

Processed only to generate your specific petition draft
Not shared with AI providers for model improvement
Not used in aggregate for AI training without anonymization and consent
Deleted according to our data retention policy (see Privacy Policy)

Anonymized Data Usage (Optional Opt-In):

With your explicit consent, we may use anonymized, aggregated data to:

Improve AI prompt engineering for better petition quality
Identify common petition weaknesses to enhance guidance
Develop statistical insights on successful petition patterns

You can opt out at any time through your profile settings.

6. AI Limitations & Disclaimers

⚠️ Important: AI is Not Perfect

While our AI systems are sophisticated, they have limitations:

Hallucinations: AI may occasionally generate false or misleading information
Outdated Knowledge: AI models have knowledge cutoffs and may not reflect latest USCIS policies
Context Misunderstanding: AI may misinterpret complex or nuanced situations
No Legal Expertise: AI is not a replacement for licensed immigration attorneys
Bias Risks: Despite mitigation, AI may exhibit subtle biases

Your Responsibilities:

✅ Verify all AI-generated content for accuracy and truthfulness
✅ Consult an immigration attorney for legal advice before submission
✅ Report errors if AI generates incorrect or biased content
✅ Provide accurate input – AI output quality depends on your input
✅ Review USCIS requirements independently to ensure compliance

7. User Rights Regarding AI

Under GDPR, EU AI Act, and ethical AI principles, you have the following rights:

Right to Human Review: Request human expert review of AI-generated content
Right to Opt-Out: Decline AI-generated drafts and use manual services
Right to Explanation: Understand how AI arrived at specific recommendations
Right to Object: Challenge AI output you believe is incorrect or biased
Right to Data Deletion: Request deletion of AI processing logs
Right to Portability: Export AI-generated drafts in machine-readable format

To exercise these rights, contact customer-support@quickpetition.com with subject "AI Rights Request".

8. AI Security & Responsible Deployment

Security Measures:

🔐 Encrypted API Calls: All data sent to AI providers is encrypted in transit
🔐 No Third-Party Training: API terms prevent providers from training on your data
🔐 Access Controls: Only authorized systems can trigger AI generation
🔐 Rate Limiting: Prevents abuse and unauthorized AI usage
🔐 Audit Logs: All AI interactions logged for accountability (90 days)

Responsible AI Deployment:

✅ Pre-deployment testing for quality and bias
✅ Ongoing monitoring of AI output accuracy
✅ User feedback integration for continuous improvement
✅ Incident response plan for AI failures or errors
✅ Regular updates to align with USCIS policy changes

9. Reporting AI Issues

If you encounter any of the following AI-related issues, please report them immediately:

❌ AI-generated false or misleading information (hallucinations)
❌ Biased content based on nationality, gender, race, or religion
❌ Inappropriate, offensive, or unprofessional language
❌ Privacy violations (AI exposing others' data)
❌ Security concerns with AI processing

Report AI Issues:

Email: customer-support@quickpetition.com

Subject: "AI Issue Report - [Brief Description]"

10. Changes to AI Systems

We may update our AI systems, models, or providers from time to time. We will notify you of material changes by:

Posting updates on this page with a "Last Updated" date
Sending email notifications for significant changes
Displaying in-app notices when AI behavior changes

Last Updated: February 2, 2026