AI can make workplaces safer by improving misconduct reporting process

• Key Insight: Learn how AI can restructure early reporting to restore trust and consistency.
• What's at Stake: Unchecked reporting failures expose organizations to legal, reputational, and operational risk.
• Supporting Data: Over 60% of workplace misconduct incidents reportedly go unreported, per Gartner.
• Source: Bullets generated by AI with editorial review

Employees don't trust leaders to look after them when they're being mistreated at work, but AI might be the key to fixing that.     

Over 60% of workplace misconduct — including harassment, discrimination and violence — goes unreported, and fewer than 15% of sexual harassment victims file formal claims with their employer, according to data from management consulting firm Gartner. Incorporating more technology and AI in the early phases of reporting could play a critical role in keeping workplaces accountable, and most importantly, their employees safe. 

"Most employers are dealing with the same core problem [which is that] reporting is not a clean, structured process," says Jared Pope, employment attorney and founder and CEO of workplace misconduct solution Work Shield. "It's fragmented, inconsistent and employees often do not trust it. AI can help in a very practical way if it is used correctly."

Read more: What the end of EEOC's harassment guide means for employers

In many cases, information arrives incomplete, gets routed to the wrong person and response timelines slip, leading to documentation of the situation being poorly handled and creating space for risk to grow, Pope explains. AI can eliminate much of those pain points by managing employee intake, organizing and keeping track of important information and asking better and smarter follow-up questions to ensure that essential elements aren't left out. 

"It can make the process easier, capture details in a consistent format, reduce back-and-forth, and flag patterns that humans often miss," Pope says. "The value of AI is structure, speed and visibility." 

Preserving the human element

However, the goal shouldn't be to remove people from the process entirely, Pope maintains, especially when employees are already distrustful. In fact, 22% of employees who witness harassment never report it, and among those who do, 38% say they were dissatisfied with their employer's response. Fear of retaliation is also one of the biggest challenges for victims and bystanders, making confidentiality key — something AI has struggled with in the past. While AI can be useful, there is still immense value to keeping leaders at the forefront.  

"An investigation requires human judgment, credibility assessment and careful handling of sensitive information — those are not tasks you want delegated to a tool," Pope says. "The best systems don't try to replace HR, they give them leverage." 

Read more: SpaceX accused of sexual harassment as fight with ex-employees intensifies

How to balance AI use

The right approach is to leverage AI only for efficiency. In practice, that means limiting AI to the administrative tasks that bog leaders down, and creating hard boundaries for its role in the rest of the process. Everything that has to do with investigations, credibility assessments and final decisions must remain human-led, Pope says. 

This not only creates a safe space for employees, but ensures that organizations have a good understanding of who made decisions, how they were made, and why the process was fair — something that's difficult if it's unclear which parts were automated versus human-evaluated.

Ultimately, only trained investigators or attorneys should be accountable for the findings and outcomes of a misconduct claim in order to protect both the employee and a business, says Pope.

Read more: Army Cadet to CEO: How this executive aims to end workplace harassment

"Misconduct reporting is not just data entry," he says. "It's people sharing difficult experiences, sometimes involving power dynamics, retaliation concerns or even serious allegations. AI can support, but it cannot assess credibility, read nuance or make judgment calls."

As AI's use cases grow, Pope hopes to see it make a positive impact on employees' physical and psychological safety. Improving these areas of the employee experience is the only way to build a happy, long-lasting workforce. 

"[Leveraging AI] is about building a process that holds up under pressure," Pope says. "When employers modernize reporting and investigations, they protect employees and they protect the business, and those two outcomes should not be in conflict, they should be directly connected."