🛡️ Built a Tool to Detect PDF Tampering – Would Appreciate Expert Feedback 🙏

[vivek_5505]

Hello Adobe Community 👋,

I’m a final-year computer science student and recently completed a personal project — a PDF authenticity analyzer 🧪 — designed to detect signs of tampering or unauthorized edits in PDF documents.

I tested it using two versions of the same file:
:white_heavy_check_mark: AI For Beginners.pdf (original)
:exclamation_mark: AI For Beginners (Modified).pdf (manually altered using a PDF editor)

:magnifying_glass_tilted_left: The tool performs the following checks:

• :spiral_calendar:️ Compares metadata (creation vs. modification dates)

• :locked_with_key: Identifies missing or invalid digital signatures

• 🧠 Highlights font anomalies and misalignments

• :warning: Detects unusual object layering or structure inside the PDF

:hammer_and_wrench:️ Tech Stack:

• Python with Flask (backend)

• PyMuPDF + pikepdf (for PDF analysis)

• Tailwind CSS (frontend)

:direct_hit: Goal:
To create a lightweight, accessible, web-based tool for document authenticity verification — especially helpful in 🎓 education, 👔 HR, and :balance_scale: legal environments.

👉 I’d really appreciate expert input on:

• Are there other PDF-level checks I should implement for better reliability?

• Is there a recommended way to validate digital signatures more accurately using Adobe’s APIs or SDKs?

• Any potential issues or blind spots in my current approach?

Thank you so much for your time and support! 🙌
Looking forward to your insights and suggestions.

Best regards,
Vivek Avdhesh Sharma
:books: Final Year Computer Science Student
🧠 Python & AI Enthusiast

[MikelKlink]

Unfortunately your description of the checks you execute is very unspecified. In particular "unusual object layering or structure inside the PDF" can mean anything. Thus, it's difficult to recommend any additional PDF-level tests. 

Similarly you say you "Identify missing or invalid digital signatures". If you already absolutely identify invalid signatures, how can we recommend a more accurate validation?

[vivek_5505]

Thank you again for the valuable feedback. I’d like to briefly clarify the purpose and motivation behind this project.

During my final year of college, I saw that many students tampered with certificate PDFs using free editors — changing issuer names, grades, etc. To help HR teams and institutions detect such edits, I created a simple tool to flag suspicious documents.

The main check is a comparison between the PDF’s creation and modification dates, which often indicates tampering. It also looks for font or layout inconsistencies and unexpected structural changes caused by external editing.

As a recent computer science graduate, my goal was to focus on practical detection, not advanced forensic analysis. The tool is useful for basic validation, especially in education and hiring scenarios.

I would love to explore how this can be improved further, possibly by collaborating with Adobe or learning how to integrate stronger PDF analysis using Acrobat SDKs.

Thanks again for the feedback and support.

Best regards,
Vivek Avdhesh Sharma
BSc CS Graduate | Creator of PDF Tampering Detector

[MikelKlink]

First of all, this is the Adobe community forum. Thus, most people here are not Adobe personal but users of Adobe products. So chances are slim Adobe will contact you in response to your post.

That being said, you say that "The main check is a comparison between the PDF’s creation and modification dates". This check is of very limited use only. On one hand there are a number of PDF manipulation libraries that don't update those entries. And on the other hand there are a number of use cases where the original document some organization provides is based on some template into which some text has been added in a second step, so in these cases differing times are no indication of fraud.

Your descriptions of the remaining checks are very vague, so one cannot get an impression on their quality.

If you want people to become interested in your product, you should describe your checks in more detail. Or provide access to a test installation people can check some example documents with.

[vivek_5505]

This is my project overview , this project is built for when i see this type of fraud
But i want to work with Adobe to discover more feature with Seal API 

and provide suggestion to stop and identify the tampering

[Joel Geraci]

The proper way to do this is to prevent tampering in the first place. Our Seal API allows the document to be "certified", making it tamper-evident in Acrobat.

[vivek_5505]

Thank you, Joel. I appreciate your response and recommendation.

You're absolutely right — preventing tampering at the source using certified documents like those supported by Adobe’s Seal API is ideal.

My project, however, focuses on cases where documents are already shared and lack any certification. In HR and education, PDFs often come from individuals and may have been edited. My tool helps detect such tampering by checking creation and modification dates and flagging structural changes.

As a recent graduate, I built this to solve a problem I noticed during college, when many students edited certificate PDFs. I didn’t know much about solutions like the Seal API at the time, but I wanted to do something useful.

Now, I’m eager to learn more and explore if my tool could be improved or integrated with Adobe’s systems.

Best regards,
Vivek Avdhesh Sharma
Recent CS Graduate | PDF Authenticity Analyzer Creator