Is Marksy Safe? Addressing Academic Dishonesty and AI Data Privacy Concerns

If you're an IB student considering using Marksy to grade your Internal Assessments, Extended Essays, or TOK work, you might have legitimate concerns about data privacy and academic integrity. This comprehensive guide addresses the most common questions and misconceptions about using AI grading tools like Marksy, explaining exactly how the technology works and why your academic work remains completely safe.

Understanding the Concerns: What Students Are Worried About

We've heard the concerns from IB students across various platforms, and they generally fall into three main categories:

"Will my work be flagged as AI-generated if I use Marksy?"
"Does the AI train on my work and store my data?"
"Am I committing academic dishonesty by using an AI grading tool?"

Let's address each of these concerns with clear, evidence-based explanations.

🎯

Struggling with IB Assessments?

Get instant, detailed feedback on your work with AI that understands IB criteria.

Start Free Trial View Subjects

How AI Detection Actually Works (And Why Marksy Won't Flag Your Work)

One of the most common misconceptions is that if you upload your work to any AI tool, it will somehow be "stored" and later detected as AI-generated. This fundamentally misunderstands how AI detection technology works.

The Reality of AI Detection

AI detection tools (like TurnItIn's AI detector, GPTZero, or Copyleaks) work by analyzing the statistical patterns in your writing, not by cross-referencing your work against a database of submissions. Specifically, they use a technique called perplexity analysis.

Here's what that means:

Perplexity is a measure from information theory that determines how "predictable" text is
AI-generated text tends to be more predictable (lower perplexity) because language models generate text by predicting the most likely next word
Human writing is typically more varied and less predictable (higher perplexity)
AI detectors are themselves smaller AI models trained to calculate this perplexity score

Important: AI detectors do NOT work by:

Comparing your text to a database of AI-generated content
Checking if your work has been "seen" by an AI system before
Accessing the training data or logs of AI companies like OpenAI

Why Marksy Doesn't Impact Detection

When you submit your work to Marksy:

Marksy reads your work to provide feedback based on IB criteria
Your original work remains unchanged - you wrote it, not the AI
The perplexity of your writing doesn't change just because an AI read it
No "signature" or "fingerprint" is added to your document

Think of it this way: If you email your essay to a friend for feedback, does that make your essay look like your friend wrote it? Of course not. The same principle applies when an AI tool reads your work to provide feedback.

Data Privacy: Does OpenAI Train on Your Work?

This is perhaps the most technical concern, but it's also the one where we can provide the clearest answer: No, OpenAI does not train on data submitted through the API.

Understanding the Difference: ChatGPT vs. API Usage

Many students confuse ChatGPT (the consumer product) with the OpenAI API (what developers use). Here's the critical distinction:

ChatGPT (Consumer Product):

Free and Plus users' data MAY be used for training (unless opted out)
Designed for general public use
Data retention policies favor model improvement

OpenAI API (What Marksy Uses):

Explicitly does NOT train on submitted data by default
Designed for business and developer use
Strict data privacy guarantees

Official Documentation from OpenAI

According to OpenAI's official API data usage policy (available at https://openai.com/enterprise-privacy):

"By default, we do not train on any inputs or outputs from our products for business users, including ChatGPT Business, ChatGPT Enterprise, and the API. We offer API customers a way to opt-in to share data with us, such as by providing feedback in the Playground, which we then use to improve our models. Unless they explicitly opt-in, organizations are opted out of data-sharing by default."

Marksy uses the OpenAI API and has NOT opted in to any data sharing. Your work is:

Not used for training
Not stored permanently (only temporarily for processing)
Not accessible to other users or systems

Additional Privacy Layer: OpenRouter

Marksy routes requests through OpenRouter, which provides an additional layer of privacy protection. According to OpenRouter's privacy documentation (https://openrouter.ai/docs/guides/privacy/logging):

Requests can be configured to avoid logging
When training opt-out is enabled on your OpenRouter account, they do NOT route to providers that train on data
Data retention is strictly controlled

Academic Integrity: Is Using Marksy Cheating?

Let's be absolutely clear: Using Marksy is NOT academic dishonesty. Here's why:

What Marksy Actually Does

Marksy is a feedback tool, not a writing tool. It:

Reads your work that you already wrote
Analyzes it against official IB criteria
Provides feedback on strengths and areas for improvement
Suggests how you might improve based on rubric requirements

Marksy does NOT:

Write your work for you
Generate content that you can copy-paste
Make changes to your document
Complete assignments on your behalf

How This Compares to Traditional Resources

Consider these perfectly acceptable IB study practices:

| Traditional Resource | What It Does | How Marksy Is Similar | |-------------------------|------------------|---------------------------| | Teacher feedback | Reviews your work against criteria | Marksy reviews against official IB criteria | | Peer review | Another student reads and comments | Marksy reads and provides structured comments | | IB Study guides | Explain what examiners look for | Marksy explains what each criterion requires | | Revision services (paid) | Human tutors grade your practice work | Marksy grades your practice work |

If getting feedback from a teacher, tutor, or peer is acceptable (and encouraged!), then getting feedback from an AI tool that does the same thing is equally acceptable.

What the IB Says About Using Resources

The IB's Academic Integrity policy states that students must:

Produce authentic work (work that is your own)
Acknowledge when they receive help or use resources
Not engage in malpractice (cheating, plagiarism, collusion)

Using Marksy for feedback on your own authentic work does not violate any of these principles. However, as with any resource:

The work you submit must still be your own writing
You should acknowledge that you used resources (including AI tools) if your teacher or school requires it
You must not submit AI-generated content as your own work (which Marksy doesn't provide anyway)

Debunking Common Myths

Let's address some specific myths we've seen circulating:

Myth #1: "AI companies own anything you input"

Reality: This is false for API usage. While some consumer AI tools have broad terms of service, OpenAI's API explicitly states that users retain ownership of their data. Your IA, EE, or TOK essay remains YOUR intellectual property.

Myth #2: "If you use AI tools, everything will be flagged later"

Reality: AI detectors cannot tell that an AI has "seen" your work. They can only detect if the writing style matches AI-generated patterns. Since you wrote your work, it will have your human writing patterns.

Myth #3: "Using any AI tool is academic dishonesty"

Reality: Using AI for feedback is like using a spell-checker, grammar tool, or getting feedback from a tutor. What matters is that the work itself is your own original creation.

Myth #4: "Your essay will end up in AI training data"

Reality: With proper API usage and privacy settings (which Marksy implements), your work is not included in any training datasets. This is a contractual guarantee from OpenAI for API users.

Myth #5: "Schools and IB can detect if you've used Marksy"

Reality: There is no technological way for schools or IB to detect that you received feedback from Marksy, just as they cannot detect if you received feedback from a private tutor, parent, or friend. What they can detect is if work is not your own or is AI-generated (which is different from AI-reviewed).

Real-World Comparison: How IB Teachers Use Similar Tools

It's worth noting that many IB teachers and schools are already using AI tools to help with their workload:

Grading assistants to provide faster feedback
Plagiarism detectors (which also use AI) to check work
Automated rubric generators for assessment
Data analysis tools to track student progress

If teachers can ethically use AI to help them grade more efficiently, students can ethically use AI to understand how their work would be graded.

💡

Pro Tip: Get AI-Powered Grading

Stop second-guessing your grades. Get instant feedback aligned with official IB rubrics.

Try Free Grading View Subjects

Best Practices for Using Marksy Responsibly

While Marksy is safe and ethical to use, here are some best practices to ensure you're using it responsibly:

1. Use It for Feedback, Not Replacement

Do: Use Marksy to understand where your work stands against IB criteria
Don't: Rely solely on AI feedback without also seeking human input from teachers

2. Be Transparent (If Required)

Do: If your school has policies about using AI tools, follow them
Don't: Hide your use of resources if your school asks about them

3. Maintain Authenticity

Do: Ensure all work you submit is your own original writing and research
Don't: Use any AI tool (including ChatGPT, not Marksy) to generate content you claim as your own

4. Use Multiple Resources

Do: Combine Marksy feedback with teacher feedback, peer review, and official IB resources
Don't: Depend on any single resource exclusively

5. Understand the Feedback

Do: Read and understand why Marksy gave certain feedback
Don't: Blindly follow suggestions without understanding the underlying IB criteria

Technical Safeguards in Marksy

For those interested in the technical details, here are the specific safeguards Marksy implements:

Data Handling

Encryption: All data transmitted to and from Marksy is encrypted using industry-standard TLS/SSL
Temporary Storage: Your documents are stored only as long as needed to process them (typically seconds to minutes)
Automatic Deletion: Files are automatically deleted from Marksy's servers after processing
No Permanent Logs: Marksy does not maintain permanent logs of your essay content

API Configuration

Training Opt-Out: Marksy's OpenAI API usage is configured to opt-out of all training
Zero Data Retention (OpenAI side): OpenAI retains API data for only 30 days for abuse monitoring, then permanently deletes it
No Model Fine-Tuning: Marksy does not use any custom fine-tuned models that would require training on user data

Access Controls

User Authentication: Only you can access your submissions through your account
No Data Sharing: Marksy never shares your work with third parties
Privacy by Design: The entire system is designed to minimize data exposure

Frequently Asked Questions

Can IB examiners tell if I used Marksy?

No. IB examiners evaluate your final submitted work based on its quality and adherence to IB criteria. They cannot tell what resources you used during your preparation process. They can only evaluate whether the work itself is authentic and meets the required standards.

Will using Marksy affect my AI detection score?

No. AI detection tools analyze the writing patterns in your text. Simply having an AI tool read your work does not change your writing patterns. As long as you wrote the work yourself (which you did), your AI detection score will reflect that you are the human author.

What if my school forbids using AI tools?

Some schools have blanket policies against AI tools. If your school has such a policy, we recommend:

Understanding what specifically is prohibited (usually AI-generated content, not AI feedback)
Speaking with your IB coordinator about whether feedback tools like Marksy fall under the policy
Respecting your school's policies while advocating for nuanced AI policies that distinguish between AI-generation and AI-feedback

Many schools are updating their AI policies to distinguish between AI-generated content (problematic) and AI-assisted learning (beneficial).

Is Marksy better than human feedback?

No, Marksy is designed to complement human feedback, not replace it. The ideal approach is:

Write your draft
Use Marksy for quick, criteria-based feedback
Seek teacher feedback for deeper, contextual insights
Revise based on all feedback sources
Repeat as needed

Marksy excels at providing instant, criteria-specific feedback 24/7, while human teachers provide nuanced guidance, answer questions, and help with conceptual understanding.

What happens to my work after I upload it?

Here's the exact lifecycle:

Upload: You upload your document through Marksy's secure interface
Processing: Marksy sends your text (encrypted) to OpenAI's API for analysis
Analysis: The AI model analyzes your work against IB criteria (this takes seconds)
Feedback: Results are sent back to Marksy and displayed to you
Deletion: Within 24 hours, your document is automatically deleted from Marksy's servers
OpenAI Deletion: OpenAI retains data for 30 days for abuse monitoring, then permanently deletes it (per their API policy)

Can other students see my work?

Absolutely not. Your work is private to your account. Marksy does not have any sharing features, public databases, or any mechanism by which other users could access your submissions.

What about GDPR and data protection laws?

Marksy complies with GDPR and international data protection standards:

You have the right to request deletion of your data at any time
You have the right to export your data
Data processing is limited to what's necessary for the service
Clear privacy policies explain exactly how data is used
You can delete your account and all associated data at any time

Could this change in the future?

Marksy's commitment to privacy is fundamental to the service, not an optional feature. Any changes to privacy policies would:

Be clearly communicated to all users in advance
Require explicit opt-in consent (not automatic)
Maintain backwards compatibility (existing users' data remains protected under original terms)

However, we encourage all users to periodically review our privacy policy and OpenAI's API terms to stay informed.

The Bottom Line: Marksy Is Safe, Ethical, and Effective

After addressing all these concerns, here's what you need to know:

Regarding Data Privacy: ✅ Your work is NOT used to train AI models
✅ Your data is NOT permanently stored
✅ Your work is NOT accessible to others
✅ OpenAI API explicitly does not train on your data

Regarding AI Detection: ✅ Your work will NOT be flagged as AI-generated
✅ AI detectors cannot tell that an AI read your work
✅ Only the writing style matters, not who provided feedback
✅ You wrote it, so it will read as human-written

Regarding Academic Integrity: ✅ Using Marksy for feedback is NOT cheating
✅ It's equivalent to teacher feedback, tutoring, or peer review
✅ The IB encourages using resources to improve your work
✅ You must still submit authentic work (which Marksy helps you improve)

Conclusion: Focus on Learning, Not Fear

The concerns about AI tools in education are understandable—this technology is new, and misinformation spreads quickly. However, when you understand how AI detection works, how data privacy is protected through the API, and what academic integrity actually means, it becomes clear that tools like Marksy are safe, ethical, and valuable resources for IB students.

Marksy is designed to help you:

Understand IB criteria better
Improve your work through detailed feedback
Learn what excellent IB work looks like
Succeed in achieving your academic goals

Just as students have always used study guides, tutors, and teacher feedback to improve their work, Marksy provides modern, accessible, and instant feedback based on official IB rubrics. Your work remains yours, your privacy is protected, and your academic integrity is maintained.

Ready to experience accurate, instant IB grading feedback? Try Marksy today and see how AI-powered feedback can help you achieve your best work.

Start Using Marksy →

Additional Resources

Have more questions? Contact us at [contact information] or visit our FAQ page. We're committed to transparency and helping IB students succeed ethically and effectively.