Labor & Employment Law Blog

AI content detectors are programs that examine text in real-time to determine whether AI was used in a document. There are several programs that employers can purchase to determine if the written product they receive was AI supported, such as GPTZero and GPT-2 Output Detector, but these programs use AI and are not always accurate. In fact, ChatGPT’s own program for this purpose, AI Text Classifier, was launched in early 2023 and was so ineffective that it was discontinued by July 2023. Thus, employers may be left with old-fashion human techniques to unravel this authorship mystery. While there is no sure-fire method to spot AI-assisted writing, there are certain reoccurring themes that have come to the forefront:

Use Of Language

According to Inc.com, there are seven words that show up more often in AI-produced content than in human-written texts:

• delves
• showcasing
• underscores
• comprehensive
• crucial
• intricate
• pivotal

Other sources have also indicated that the words such as “cutting-edge” and “innovative” are repeatedly used by AI, as are “firstly,” “secondly,” and “in conclusion.”  Used together, these “marker words” increase the odds that a text was created by AI, particularly texts originating from countries where English is not the native tongue. Besides use of these words, be on the lookout for repeated use of certain words or synonyms. AI often utilizes the same words to describe concepts of ideas while human writers tend to avoid repetition and employ diversity in their writing. The lack of human emotions is another dead give-away of AI writing. Words such as “like,” “cute,” “love,” and “believe” often cannot be found in AI-generated text. Another clue is how paragraphs start. AI tends to overuse the phrases, “However,” “Moreover,” “Furthermore,” and “In addition,” while human writers tend to vary their transitions and introductions. Finally, the speed of a response can indicate non-human authors. Rapid responses could be a sign of AI.

The Humanistic Approach

There is no substitute for human interaction to determine if someone is an author or if an employee truly understands the subject matter of a writing. The best way to test people is by posing hypotheticals. This personal interaction will show the depth of a person’s knowledge. The major disadvantage of this methodology is that it is time-consuming. Otherwise, when looking at a piece of writing, some clues that AI may have written a document are grammatically correct sentences with no spelling errors that lack humor and personal touches.

Why We Should Care

If AI is here to stay, why should companies care if applicants and employees use the technology? AI writing brings up many issues related to ethics, integrity, and morals. There is an importance to spotting use of AI to make sure that people are using it correctly. Companies should want to take advantage of any tool that gives them a competitive advantage, but because of the inherent inaccuracies in AI, it is critical to make sure that employees do not solely rely on AI’s work-product. The use of AI in technological fields can facilitate communication among people who are not necessarily known for their ease of communication with the written word. Helping people to elevate their written communications to express ideas more effectively is a business tool worth using with proper controls and reviews. The goal should not be to blame people for failing to fact check. Rather, the goal should be to correctly leverage the newest tool in the workplace arsenal.

Employer Takeaways

• Use of AI offers significant benefits including efficiency and cost-effectiveness, but also presents challenges like lack of originality, potential biases, and quality concerns.
• AI generated content is most effective for routine tasks such as summaries and social media post, but should be avoided for complex topics that demand nuanced human expertise.
• AI text detection is crucial for uncovering low-quality content that should not be used without editing and fact-checking.
• To achieve the best results, it is essential to combine AI-generated content with human input for fact-checking, editing, and ensuring content integrity.

  This AALRR publication is intended for informational purposes only and should not be relied upon in reaching a conclusion in a particular area of law. Applicability of the legal principles discussed may differ substantially in individual situations. Receipt of this or any other AALRR publication does not create an attorney-client relationship. The Firm is not responsible for inadvertent errors that may occur in the publishing process.

  © 2025 Atkinson, Andelson, Loya, Ruud & Romo