MANILA, PHILIPPINES — As AI-generated text becomes increasingly prevalent, educational institutions and professionals are searching for effective ways to identify it. While tools for detecting AI content are evolving, the rapid advancement of AI technology has made it possible to bypass these measures.
Although using AI for unethical purposes like automating homework is discouraged, understanding how detection systems work can help educators and users navigate this landscape responsibly.
Here are some techniques to make AI-generated text less detectable.
1. Leverage Tools Like Undetectable.ai
Undetectable.ai is a tool specifically designed to make text harder for AI detection systems to identify.
How it works:
- Visit the Undetectable.ai website.
- Sign up for an account.
- Paste the AI-generated content into the editor.
- Agree to the Terms of Service, which prohibit academic misconduct.
- Click the “Humanize” button to revise the content.
The platform can also assess whether tools like Copyleaks or ZeroGPT would flag the revised text.
2. Ask ChatGPT to Rewrite Content
Another approach is to request that ChatGPT rephrase its own output. Over several iterations, the text may become less recognizable as AI-generated.
Tips:
- Be patient, as multiple rewrites may be necessary.
- Use grammar tools to refine the final version, ensuring readability and coherence.
3. Use Paraphrasing Tools Like Quillbot
AI paraphrasing tools can help by rewording text while maintaining the original meaning.
Steps:
- Visit Quillbot.com.
- Paste your text into the editor.
- Click “Rephrase” to generate suggestions.
Quillbot also offers word and phrase replacements to further refine content.
4. Revise AI-Generated Text Manually
The most reliable way to humanize AI-generated text is through personal editing.
Suggestions:
- Break repetitive sentences into shorter ones or rearrange clauses.
- Replace commonly used words with synonyms.
- Add transitions and conversational phrases like “I think” or “you might.”
- Use contractions to make the tone more natural.
5. Use Detailed Prompts and Customized Instructions
Crafting precise prompts can make AI-generated content more authentic. For example, instead of a generic command, create detailed instructions tailored to your needs.
Example:
“Write a 150-word science report about photosynthesis as if you were a fifth-grade student.”
To make this process easier, utilize ChatGPT’s “Customized Instructions” feature to save your preferences for future interactions.
Responding to AI-Generated Content
As detection tools evolve, so will methods to evade them. Banning AI tools entirely is unlikely to succeed, as innovation tends to outpace regulation. Instead, fostering responsible AI use and educating people about its capabilities is a more sustainable approach.
