In 2017, technology is a part of our everyday experience. In a world where Amazon has replaced the corner store and Tinder has replaced the local bar, it’s not surprising that businesses are now turning to apps and artificial intelligence to assist with workplace harassment reporting and investigations.
Using artificial intelligence (AI) to identify potentially dangerous situations is nothing new. Facebook has used it to identify terrorist propaganda, Wikipedia uses it to combat internet trolls, and some even believe that AI can be a useful tool in the war against “fake news”. So what can it do for employers who want to keep an eye out for potential harassment?
Recently, some employers have turned their minds to using AI to analyze data from conversations exchanged within inter-office communication systems, such as Slack and Microsoft Teams, in order to recognize potentially harassing language. Many workplaces are already using these systems to monitor productivity, and keeping an eye out for harassing behaviour might seem like a logical next step. A bot installed on the messaging system, or on the employer’s email server, can be set up to identify communications that contain specified words or terms, and flag them or block the content.
In terms of harassment reporting, the TTC recently announced that it is developing an app to report sexual harassment and assaults on the subway. An app called Callisto is already being used in colleges to encourage the reporting of sexual assaults on campus, and another app – StopIt – can be used to allow employees to anonymously report bullying and harassment in the workplace.
There are some potential pitfalls of depending on AI and apps to deal with workplace harassment. One obvious issue is that human interactions are complex; no list of defined terms – however long – will be able to capture every harassing message exchanged within a workplace. There is danger in over-relying on such a system to identify whether a staff member is being harassed.
The flipside of this is that many innocuous communications will likely be captured as well. Terms that can be harassing in one context can be completely appropriate in another, and having work-related emails repeatedly flagged for potential harassment could lend an unpleasant “big brother” atmosphere to a workplace. Employers also might want to consider whether the resources that would be expended reviewing these innocuous communications could be better spent on other anti-harassment initiatives.
With respect to the use of apps to report harassment, this makes perfect sense in situations where the victim might not be able to seek out immediate assistance from a person they trust, such as on the bus or subway. When it comes to using these apps as a means of anonymously reporting harassment in the workplace, however, there may be difficulties.
While employees who fear reprisal might be more comfortable reporting bullying and harassment anonymously through an app, anyone who has investigated workplace harassment complaints will attest that there are challenges associated with anonymous complaints.
First, it is rare that the initial complaint document contains all the information an investigator needs to evaluate the complaint; this information is usually obtained through a detailed interview with the complainant, and often one or more follow-up interviews. Second, most incidents of bullying and harassment arise through one-on-one exchanges, and accordingly the investigator must carefully assess two different stories. The credibility of the parties can be a key part of this, and information gathered during a face-to-face interview can be vital to an evaluation of credibility. Finally, in the interests of fairness, the respondent should be given a chance to fully respond to the allegations against them. Not knowing whom they are alleged to have harassed or bullied can present challenges in defending themselves.
Technology can be a valuable device for the identification, reporting, and investigation of harassment allegations, but it is also important for both investigators and employers to be aware of its limitations. If potential concerns are raised anonymously through an app, or through identification of harassing language in inter-office communications, employers should consider whether a more in-depth review – such as a workplace assessment – would be a useful tool to better gauge the full extent of the problem.