Code Plagiarism vs MOSS — Why Stanford's 1994 Algorithm Isn't Enough Anymore

Detecting cases of code plagiarism with results from billions of sources

Measure of Software Similarity, or MOSS, is a piece of software used by teachers and publishers to find software plagiarism. This software was developed by Stanford, and has been the primary tool for checking code plagiarism. Codequiry is a similar program to the MOSS code. While MOSS was the first, Codequiry was created to improve on the weakness of moss program checker. Both were created to stop the prevalence in students to copying code, instead of creating new code. Students were not learning. The students that teachers were giving the higher grades to turned out to be those who could find the best code via the internet. This has changed with the creation and use of Codequiry.

Ease of Use

One of the ways that Codequiry has improved on MOSS plagiarism checker is by being made to be easier to use. Teachers and instructors do not need to spend a lot of time just going through a piece of software. That is one of the reasons that Codequiry has become as popular as it has. Less time with software allows more time with the students.

Result documentation

Sometimes a student may actually have just written software that is similar enough to set off the software. MOSS code checker will not know this. Codequiry saves results so that if someone has questions, they can review what is found. With software programming being such that there can be valid similarities, Students need to be able to discuss their process.

Checking Code for Plagiarism from the Web

One of the ways that any software is able to keep up is utilizing the internet. Checking via the web allows for finding newer examples that may not be available, even in an update. Instead of relying on programmers finding examples and then putting it in an update, internal artificial intelligence can search just as well. This allows for instructors to worry about the check, not updating. Moss dosen’t have a online checker which provides a massive loophole for code cheaters to use external web sources. Codequiry engine was designed to be able to detect up to 11 billion different web sources for evidence of plagiarism. For checking code similarity its critical to have both web and peer check to have a full picture of the samples.

An Advanced Code Plagiarism Detection Engine

Codequiry’s code engine checker is customizable in a way other code checkers aren’t. Users have the power to choose between a list of different type of checks based on the situation. Measure of software similarity doesn’t have a variance of checks in the way Codequiry does. For example, if a person wants to just examine peer samples for code similarity they can. If they need to go more in depth in checking for external sources the engine gives any user that specific option. The advanced engine gives more control and precision that is more specific to the user’s situation.

The AI-Generated Code Problem MOSS Was Never Designed to Solve

MOSS was built in 1994, when the worst-case copy was a student emailing another student a .c file. In 2026, the dominant copying mechanism is generative AI — ChatGPT, Claude, Gemini, GitHub Copilot — producing fluent, idiomatic code in milliseconds. Stanford's MOSS plagiarism checker has no concept of AI-generated code. It can only flag overlap between submissions, so if every student in a class prompts ChatGPT with the same homework, MOSS will report low similarity and everyone passes. Codequiry's AI code detector runs alongside the similarity check on every submission and produces a confidence score for ChatGPT, Claude, Gemini, and Copilot output — closing the loophole that has rendered MOSS-alone workflows obsolete.

Codequiry vs MOSS — Side-by-Side Comparison

MOSS still has its place as a baseline algorithm — and that's exactly why Codequiry includes it as one of 20+ detection engines, not the entire product. You get everything MOSS does, plus everything it can't.

When to Choose Codequiry Over MOSS

If you're a CS educator running a class of more than ten students, a coding bootcamp grading hundreds of submissions per cohort, or an enterprise team auditing code provenance for legal or M&A purposes, the gap is decisive:

• Use MOSS if you have a single small class, no AI-cheating risk, and the time to interpret raw text output by hand.
• Use Codequiry if you need a complete code plagiarism checker with web matching, AI detection, defensible evidence, and a workflow that fits your LMS — without rebuilding tooling around a 30-year-old academic project.