CrossPlag is one of the newer AI-detection platforms on the market. In contrast to other platforms we’ve reviewed, such as ZeroGPT, they not only provide AI content tools but also offer plagiarism detection features. This review will examine the advantages and disadvantages of CrossPlag.
The use of AI in writing has snowballed due to recent advancements in artificial intelligence. ChatGPT and similar artificial intelligence technologies can quickly produce well-written, polished, and unique content that is often difficult to distinguish from human-authored text. CrossPlag is one tool for being able to detect AI-generated content, unless the writer uses an AI humanization tool like Undetectable.ai.
However, concerns regarding content originality still exist. Additionally, search engines favor original and valuable content, often missing in AI-generated text. For this reason, many writers and marketers need to use AI detection bypass tools to disguise their use of AI technology.
Exploring CrossPlag’s AI content identification tool may be advantageous, which accurately identifies AI-generated content, even when it appears to be written by humans.
This article provides a comprehensive review of CrossPlag’s AI content identification tool. In this review, I conducted thorough tests to assess the tool’s effectiveness using human text, AI-generated text, and humanized inputs.
CrossPlag: Human-Generated Content & Inputs
Like in all of our previous reviews, The first test we ran for this Crossplag review was two human-written pieces from different parts of the internet and our own writings.
Test 1: Wikipedia
This input was taken from Wikipedia’s “Golden Age of Piracy“.
A.I. Detection Results:
As shown in the image above, CrossPlag was successfully able to identify the Wikipedia article was in fact, written by a human.
Test 2: Student Written
The second input was submitted by a student from a local university.
A.I. Detection Results:
As you can see, CrossPlag falsely labeled this human-written input as 46% AI. The Tag states that “This text is co-written by both a human and an AI“
CrossPlag: AI-Generated Content & Inputs
The second group of tests we ran for this CrossPlag review was two AI-written pieces from the AI generation platform ChatGPT.
Test 1: ChatGPT – Justice in America (Essay)
The first input we put into CrossPlag for the AI-generated piece was an essay about Justice in America.
Test 2: ChatGPT (Email)
The second input we put into CrossPlag for the AI-generated piece was an email asking a friend to cancel their credit card due to fraud.
CrossPlags algorithm was more confident of the likelihood that the email was written by AI than it was of the essay.
CrossPlag: AI-Generated Content Humanized by a third-party paraphrase to attempt to bypass AI detection
Several companies in the marketplace are attempting to humanize AI-generated text to help marketers rank their content on Google, or in some cases, even trying to help students with attempts at academic dishonesty. These tools claim to help companies either bypass AI detection or make AI-generated content undetectable.
Test 1: Quilbot – Justice in America (Essay)
In the following test, we ran our ChatGPT-generated essay that failed in our last test through the popular paraphrase and text humanizer Quilbot. CrossPlag claims it can accurately identify content spinners such as Quilbot. Let’s put that to the test.
As you can see, the results from Quilbot’s paraphrasing have successfully stumped CrossPlag’s AI content detection system.
Test 2: Undetectable – Justice in America (Essay)
We ran the AI-flagged ChatGPT-generated essay in the second test through the more advanced Undetectable text humanizer.
The results show that Undetectable could convert the AI-generated text into an output flagged as 100% original and human!
Price: CrossPlag is not the most accurate tool we have tested in the marketplace, but it is free for research-constrained people.
Accuracy: CrossPlag is not one of if not the most accurate tools that we have tested in the marketplace. It did not appear to be able to identify basic content spinners such as Quilbot.
In conclusion, CrossPlag seems able to identify ChatGPT sentences accurately, at the expense of occasional false positives. In fairness to CrossPlag, they are in their beta, and I’m sure the product will improve over time.