AI Transparency

PaperDesk uses artificial intelligence (AI) to help educators create educational content more efficiently. Our AI tools can: • Generate mock tests and question papers based on user-specified parameters (subject, topic, difficulty) • Create educational challenges in multiple formats (instant feedback, scorecard, leaderboard) • Validate and enhance question quality • Provide performance analytics and insights The AI is designed to be a productivity tool for educators — it handles repetitive content generation tasks so teachers can focus on instruction and student interaction.

PaperDesk's AI is designed to assist, not replace, educators. We firmly believe that: • Teachers bring irreplaceable expertise, judgment, and empathy to education • AI-generated content requires human review and validation • The most effective learning environments combine technology with skilled educators • Classroom instruction, student mentoring, and personalized guidance cannot be automated PaperDesk is a tool in the educator's toolkit, not a substitute for qualified teachers.

All content generated by PaperDesk's AI should be reviewed and verified by qualified educators before use. Specifically, we recommend: • Review questions for factual accuracy and curriculum alignment • Verify that difficulty levels match student capabilities • Check for inappropriate, biased, or misleading content • Make necessary edits before distributing to students • Periodically audit AI-generated content for quality Teachers are ultimately responsible for the educational content used in their classrooms.

While PaperDesk's AI is powerful, it has important limitations: • The AI can make mistakes — it may generate incorrect answers, irrelevant questions, or misleading explanations • The AI's knowledge is based on its training data and may not reflect the latest curriculum changes • The AI may exhibit biases present in its training data • The AI cannot understand context with the same depth as a human educator • Complex, multi-step reasoning problems may not be handled reliably We continuously work to improve our AI models, but we cannot guarantee perfect accuracy.

PaperDesk is committed to responsible AI practices: • Privacy — Your data is not used to train or improve AI models without your consent • Transparency — We clearly indicate when content is AI-generated • Human Oversight — We recommend human review of all AI outputs • Fairness — We strive to minimize bias in AI-generated content • Accountability — We take responsibility for our AI systems and continuously improve them We encourage users to report any concerns about AI-generated content to support@thepaperdesk.com.

PaperDesk uses the following AI technologies: • Anthropic (Claude) — Used for question generation and content creation • OpenAI — Used for question validation, quality assurance, and fallback generation These models are accessed through their respective APIs. We do not run AI models locally. Prompts and content sent to these APIs are processed in accordance with our agreements with Anthropic and OpenAI, which prohibit them from using your data for model training.

If you encounter AI-generated content that is: • Factually incorrect • Inappropriate or offensive • Biased or discriminatory • Otherwise problematic Please report it to us at support@thepaperdesk.com with the subject "AI Content Issue." Include details about the content, how it was generated, and why you believe it needs attention. We take all reports seriously and will investigate promptly.