By Karen Carrera
Artificial intelligence has arrived in the workplace, and it is increasingly appearing in an unexpected place: employee complaints.
Human Resources professionals, attorneys, and workplace investigators are reporting a growing number of complaints, witness statements, and requests for investigation that were clearly drafted with the assistance of AI tools. Some complaints are polished, detailed, and legally sophisticated. Others contain language that seems strangely formal, overly emotional, or disconnected from the employee’s normal communication style. As a result, many employers and attorneys have become suspicious whenever they encounter an employee complaint that appears AI-generated.
But should they be?
As workplace investigators, we may be asking the wrong question. Rather than focusing on whether AI was used, we must nevertheless should focus on whether the allegations warrant investigation.
The rise of AI-generated complaints presents both opportunities and challenges. Understanding the distinction is becoming an increasingly important investigative skill.
The Stigma Surrounding AI-Generated Complaints
Many management-side attorneys, HR professionals, and supervisors react negatively when they discover that an employee used AI to draft a complaint.
The typical reactions sound something like this:
“This doesn’t sound like the employee wrote it.”
“The complaint is full of legal terminology.”
“Someone must have fed facts into ChatGPT.”
“The employee probably exaggerated everything.”
“The complaint seems manufactured.”
These reactions are understandable, but they may also reflect assumptions that are not supported by evidence.
The reality is that AI did not create the underlying facts. At most, AI assisted in communicating them.
A complaint drafted with AI is not fundamentally different from a complaint drafted with the help of an attorney, a union representative, a trusted friend, a spouse, or a professional writing coach. Employees have long sought assistance in organizing their thoughts and expressing their concerns. AI is simply the newest tool available for that purpose.
Does AI Help Level the Playing Field?
In many situations, the answer is yes.
Investigators routinely encounter employees who are intelligent, credible, and truthful but struggle to communicate effectively in writing.
Some employees may:
- Have limited English proficiency.
- Have learning disabilities.
- Be uncomfortable with written communication.
- Become emotional when writing about workplace experiences.
- Have difficulty organizing events chronologically.
- Feel intimidated by management or HR.
Traditionally, those employees often submitted complaints that were incomplete, disorganized, difficult to follow, or lacking detail. Many of them never put their complaints in writing at all. They relied on a verbal complaint that may or may not have been taken seriously because it was not in writing.
AI can help bridge those communication gaps.
An employee who enters a series of factual notes into an AI system may receive back a more coherent narrative that better explains what happened. The facts may be the same. Only the presentation has changed.
From an access-to-process perspective, that may be a positive development.
Employees should not receive less consideration merely because they lack professional writing skills.
Why Employers and Attorneys Often Distrust AI-Generated Complaints
Despite these benefits, there are also legitimate reasons for concern.
AI systems are designed to generate persuasive and comprehensive narratives. In doing so, they may introduce language that the employee never would have used independently.
Investigators occasionally encounter complaints containing phrases such as:
- “hostile work environment”
- “retaliatory conduct”
- “systemic discrimination”
- “abuse of authority”
- “pattern and practice”
The employee may not fully understand these terms. They may have copied language suggested by AI.
As a result, some complaints become inflated. Relatively ordinary workplace disputes may suddenly be framed using legal concepts that do not accurately reflect the underlying facts.
This phenomenon can make investigators wary.
However, the concern should not be that the complaint is AI-generated. The concern should be whether the legal conclusions embedded within the complaint are supported by actual evidence.
After all, investigators are not charged with determining whether employees accurately understand employment law. They are charged with determining what happened.
AI Does Not Necessarily Mean False
One of the more troubling assumptions emerging in workplaces is the belief that an AI-generated complaint is inherently less credible than a traditionally drafted complaint.
That assumption is flawed.
The use of AI tells us virtually nothing about whether the allegations are true.
An employee may use AI to write a completely truthful complaint.
Another employee may write a completely fabricated complaint without any AI assistance whatsoever.
Credibility is determined by evidence, corroboration, consistency, motive, plausibility, and witness testimony, not by the drafting tool used to create the complaint.
Investigators should therefore resist the temptation to draw adverse inferences merely because a complaint appears professionally written or AI-assisted.
Using AI is not evidence of dishonesty.
How Can Investigators Distinguish Fact from Fiction?
This is where investigative fundamentals matter.
The answer to AI-generated complaints is the same answer that has always existed: investigate the facts.
When reviewing complaints, investigators should separate factual assertions from conclusions.
Consider the following statement:
“My supervisor created a hostile work environment through ongoing discriminatory conduct.”
That sentence contains conclusions.
The investigator’s task is to identify the underlying facts:
- What specifically was said?
- Who said it?
- When was it said?
- Who witnessed it?
- How often did it occur?
- Was there documentation?
- Are there corroborating witnesses?
- Are there contradictory facts?
Similarly, if a complaint states:
“Management retaliated against me after I reported misconduct.”
The key issue is not whether the employee used the word “retaliation.”
The key issue is whether there is evidence that:
- The employee engaged in protected activity.
- Management knew about it.
- An adverse action occurred.
- A causal relationship exists.
The investigative process remains unchanged.
Watch for AI Hallmarks Without Overreacting
Some complaints may contain indicators that AI was involved. These can include:
- Excessively formal language.
- Overly detailed legal terminology.
- Repetitive phrasing.
- Lengthy narrative sections with minimal factual specificity.
- Generalized conclusions unsupported by examples.
These features may tell you something about how the complaint was drafted. They tell you almost nothing about whether the allegations are true. Investigators should therefore avoid treating AI indicators as red flags for deception. Instead, they should view them as signals that additional clarification may be needed during interviews.
The Better Question: What Did the Employee Actually Experience?
When an employee submits an AI-generated complaint, one of the most effective interview techniques is to return to the employee’s own words.
For example:
“Tell me what happened.”
“Walk me through the events from the beginning.”
“What specifically did you experience?”
“What do you mean when you say ‘hostile work environment’?”
“Can you give me examples?”
These questions help separate the employee’s actual observations from legal conclusions or AI-generated language.
Often, investigators discover that the employee’s firsthand account is far more useful and nuanced than the written complaint itself.
Should Employers Investigate AI-Generated Complaints?
In my view, absolutely.
The mere fact that a complaint was drafted with AI should never determine whether an investigation occurs.
Employers have long had obligations to respond appropriately to allegations of harassment, discrimination, retaliation, workplace violence, bullying, ethics violations, and other forms of misconduct. Those obligations do not disappear because technology assisted in drafting the complaint.
The threshold question should remain:
Does the complaint raise allegations that, if true, would violate policy or warrant further inquiry?
If the answer is yes, the matter should be assessed and investigated using the same standards applied to any other complaint.
Rejecting complaints solely because they appear AI-generated of course creates significant risk. It may cause organizations to overlook legitimate concerns and undermine employee trust in reporting systems.
Conclusion
AI-generated employee complaints are likely here to stay. As investigators, we should be cautious about treating them as inherently good or inherently bad.
AI can help employees communicate more effectively, organize facts, and articulate concerns that they may otherwise struggle to express. At the same time, AI can introduce legal conclusions, exaggerate language, and create narratives that require scrutiny.
The critical point is that AI does not determine truthfulness. Evidence does.
The investigator’s role has not changed. Separate facts from conclusions. Test allegations against the evidence. Assess the credibility of each witness and party by using the credibility factors. Corroborate where possible. Follow the facts wherever they lead.
In the end, the question is not whether a complaint was written by a human, by AI, or by a combination of both.
The question is the same one investigators have always been tasked with answering:
What actually happened, and what does the evidence show?
Contact RPLG if you need a prompt, thorough, and unbiased investigation. We can help. Call us at 415-504-1856 or email us at KCarrera@publiclawgroup.com.