Turnitin’s AI detector is one of those detectors that has made lives a living hell for many of the students out there. Just like any other AI detector it is riddled with false-positives, and what is more is that many Universities around the world trust it blindly.
Although, some universities have woken up and banned the use of AI detectors but there are many who haven’t done anything yet.
Their AI detector wasn’t so bad before the Dec 7th, 2023 update. It was not that sensitive and therefore rarely had any false-positives. Even if it flagged, a simple rephrase with Quillbot or any other tool fixed it.
However, it became completely different after it and now any kind of simple rephrase will not work. Even if you do manual rephrasing of the text it won’t work.
So, lets dive into what are the current techniques that could help people in bypassing Turnitin’s current version of AI detector.
Ways to bypass Turnitin’s AI detector
1. Play with the punctuations and don’t use Grammarly
Since the rules of the English language are not very discrete you can use it to your advantage and play with the punctuations.
Here are some of the confusion areas which you could easily exploit it to your advantage:
- Oxford/serial comma usage
- Semicolon vs. colon usage
- Comma placement in complex sentences
- Quotation mark placement with other punctuation
- Hyphen vs. en dash vs. em dash usage
One common mistake which people make is to use Grammarly after actively employing tricks to bypass Turnitin. Please don’t use Grammarly or any kind of grammar checker as it will ruin your objectives.
Grammar checkers like Grammarly will not only correct your grammar mistakes it will also use a particular grammar style and remove any inconsistencies that you have employed purposely. This is enough to trigger Turnitin because LLMs write very fluent English following a particular grammar style.
2. Modify your writing style
You need to edit your writing style, if you often get flagged by AI. Some people have this issue where whatever they write, they get flagged by AI. Usually these people have perfect grammar, and are very fluent in English.
Read the works of editors of major publications like Wikipedia, Nature, Conversation, etc. The creators of AI detectors are usually aware of such sites and they know their detectors must not flag their writing styles as it would lead to heaps of bad press. So, you could read through them and see if you can incorporate their writing style into your own.
3. Use an AI text humanizer
AI text humanizers are tools that rewrite text for a particular reason i.e. to bypass AI detectors like Turnitin. They use a variety of techniques to do it but they all use LLMs. So, the point is if you absolutely hate LLMs for writing then you can avoid this method. However, if you do have the money then this is literally the easiest path.
Many such AI text humanizer like Deceptioner are regularly updated in order to be a step ahead of these AI detectors.
4. Follow a consistent citation style
You need to select one from the following citation styles like APA, MLA, Chicago, and Harvard styles. Using citations often throw AI detectors off the track and they will pass you as human. However, don’t overdo it as it will look weird to your professor and they might scrutinize it further. Proper citation indicates that you can combine your personal thoughts with information from research.
5. Use active voice
This is one of the more subtle ones. Using active voice do seem to bypass Turnitin’s AI detectors. However, the problem with this approach is that academic writing is very bland and you cannot apply it all the time. The point is that you need to use it whenever you can.
Frequently Asked Questions
Q1. How accurate is Turnitin’s AI detection?
Despite the claims on their website, Turnitin’s AI detection is grossly inaccurate. They often flag completely hand-written content. This is why a handful of universities have completely banned the use of AI detector software like Turnitin.
Q2. Can Turnitin detect Quillbot 2024?
Yes, obviously it can detect Quillbot because of two simple reasons. The most important one is that Quillbot’s paraphrasing doesn’t targets to bypass AI detection. Second one, is that Quillbot is way too popular and those who made Turnitin’s AI detection would have considered this and implemented guardrails against it.
Q3. Is ChatGPT detectable by Turnitin?
Yes, it can easily detect ChatGPT if you don’t use any tricks in order to bypass it. ChatGPT generally outputs text that have very low burstiness or perplexity scores.
Q4. Is GPTZero better than Turnitin?
No, it is very similar or even worse than Turnitin. But one thing you can take out of it is that if you can bypass GPTZero then that text will get bypassed for Turnitin as well.
Q5. Can Turnitin detect if I paraphrase it?
You can do it only if you paraphrase keeping in mind that you need to bypass Turnitin. The best way to do it is manually. However, you can use AI text humanizers like Deceptioner.
The Bottom Line
Turnitin is good when it comes to plagiarism but for AI detection it is hot garbage. One thing which I would like to add is that AI detectors are still in their early stages. I think it will get better overtime but the thing is those anti-AI detectors will get better too. So, basically it is a cat and mouse game and it will get interesting overtime.
