Recently, a marketing colleague asked if she should use AI detection tools to determine whether the submitted content was original or if it was developed by AI (artificial intelligence)?
It’s a very relevant question, considering how impactful AI tools have been since ChatGPT’s release in 2022.
My answer is probably not what she was hoping for because I suggested that rather than focusing on the content origin, the focus should be on the risk of plagiarism. This approach can provide a more comprehensive understanding of the content’s authenticity.
What is an AI Detection Tool?
Ironically, AI text detection tools use software that relies on AI to sniff out content generated by AI. I find this ironic because it’s like having an accounts payable person sniff out fake invoices they’ve created for their embezzlement scheme.
Jeff Bullas states, “AI detectors can be used for various purposes, including fraud detection, content moderation, spam filtering, malware detection, and identifying AI-generated content.”
How to bypass AI detection tools.
When leveraging AI tools to generate content, here are some best practices to help your content avoid getting caught in the AI detection tool web.
Start with an outline.
When interviewing clients for the content they want me to create, I first provide them with a list of questions. That list helps me to organize the final piece. It also drives the natural flow of the conversation, which I then use in the AI tools for follow-up prompts that surface naturally during the conversation.
Use smaller prompt blocks.
Rather than using complex prompt blocks that generate a large portion of text, consider using smaller blocks that generate a more natural content flow. Then, review and edit each segment before moving on to the next segment to develop a more natural follow-up prompt.
Remove redundant phrases.
Once your piece is complete, run it back through the AI tool to identify and modify redundant phrases. You can often do this within the original AI tool or use a secondary tool to assist. For example, if I create a blog post using ChatGPT, I will then use Grammarly to help me refine and edit the content.
Check your product’s settings.
In many AI content generation tools, you can modify the default settings to maximize brand effectiveness and authenticity. In ChatGPT, for example, go to Settings > Personalization > Custom Instructions and complete the information blocks provided to customize content for your brand.
Inject the brand’s personality.
There’s nothing more effective than injecting your brand’s voice, tone, and personality into AI-generated content. This helps make the content more personalized, engaging, and less auto generated.
AI Detection Tools
There are dozens of AI generation detection tools, including tools created by OpenAI. However, they recently pulled their AI Text Classifier product, which was designed to identify AI-generated content, due to its low accuracy.
If you’re considering an AI detection product, here’s a short list to check out:
The Human’s Responsibilities
Whether generating content using the human brain, AI tools, or a combination of both, it’s important to review the content’s accuracy. In a recent copywriting course I taught, I stressed to participants that the human’s writing and content creation responsibilities include:
- Ensuring accuracy
- Checking links
- Leveraging brand voice
- Sounding conversational
- Testing
No matter what you choose to create content today, remember AI-generated content is not going away any time soon. In fact, Gartner recently claimed that by 2025, 30 percent of outbound marketing messages from large organizations will be synthetically generated.
What’s your biggest concern about AI-generated content and its use in accounting marketing?