A new trend is appearing in HR offices and legal departments across the country: Employee complaints and legal documents that seem professionally written but show clear signs of being created with generative AI.
I’m not referring to employees getting legal advice from ChatGPT (which, as of this week, ChatGPT itself says it shouldn’t be used for). I’m talking about people using ChatGPT and similar tools to write their harassment complaints, discrimination charges, and demand letters.
Technology has made complex legal language easier to understand, allowing employees to use these tools to more effectively address workplace issues.
This isn’t necessarily a problem. Employees are entitled to ask for help expressing their concerns. However, employers should be aware of this trend and know how to respond effectively.
AI-generated complaints often have similar features, such as:
- Overly formal language that doesn’t match how the employee usually communicates. For example, an employee who typically writes short, informal emails suddenly submits a long complaint using precise legal terms.
- Structured legal arguments that cite specific laws, regulations, or legal standards the employee likely wouldn’t know. For example, a retail worker’s complaint refers to Title VII rules as precisely as an employment lawyer would.
- Using bold fonts (especially for parts of sentences), subheadings, bullet points (please disregard this post), emojis, and other elements that most people would not typically use when writing complaints.
- Every sentence is perfectly grammatically correct and well-formatted. There are no typos, no informal language, and no emotional expressions that might be expected from someone genuinely upset about workplace treatment.
- The complaint uses complex legal language to describe general violations, but it lacks clear details such as specific dates, times, and names of witnesses.
- Long paragraphs discuss every possible legal theory. The complaint lists hostile work environment, retaliation, failure to accommodate, and constructive discharge, even if only one of these actually fits the facts.
Now that you can recognize a possible AI complaint, let’s clarify what this actually means—and what it doesn’t.
An AI-assisted complaint is not automatically false or frivolous. What matters most is the actual content of the complaint, not how it was written. If an employee has a real harassment concern and uses AI to help express it, their complaint should be investigated just as seriously as if they had reported it in person.
What this means: You need to look beyond the formal language to understand the real allegations. Don’t let complex wording make you overreact. Also, don’t ignore concerns just because they sound overly legal or technical.
The real challenge is telling the difference between genuine concerns raised with the help of AI (which require your usual investigation and response) and suspicious concerns that are actually excuses to avoid performance issues or attempts to get a better severance package.
This is a growing trend that I expect will speed up as generative AI tools become more advanced and easier to use.
I am certain that more employees will use them to write workplace complaints, EEOC charges, demand letters, and similar documents.
Your basic responsibilities remain the same. You still need to investigate complaints, address workplace issues, and follow employment laws. The main difference is that you may now receive more clearly written and detailed documents from employees who previously had difficulty expressing their concerns.
The silver lining: Sometimes, AI-assisted complaints are clearer about the actual issues than emotional, rambling complaints. This makes it easier to identify what needs to be investigated and addressed.
The key for employers is to focus on what really matters, not just appearances.
If you’re noticing more complex employee complaints at your organization, you’re not alone. This is becoming the standard experience for many organizations.
The real issue isn’t whether employees are using AI. It’s whether you’re addressing their concerns properly, no matter how those concerns are expressed.