HR Automation

Key Points 

• In the future, Ontario employers making publicly advertised job postings will be required to disclose within the job posting if they use AI to screen, assess or select applicants for publicly advertised job posts 

• Where a job applicant participates in a video interview and sentiment analysis is used, there could be privacy and human rights issues that are triggered 

• Other jurisdictions, such as New York City, have addressed AI tools in recruitment and hiring very differently than Ontario 

My co-founder, Tommy Cooke, just wrote an informative article regarding some of the main HR automation trends that have been pervasive in the business world in 2024.  

When it comes to these trends, it is worth taking a closer look at some of the issues that could become problematic. More specifically, I would like to examine the uses of AI in the area of recruitment and hiring. 

Whether it is using AI to automate resume screening or using AI to conduct video interviewing sentiment analysis, there could be some challenges for employers. 

In particular, employers will need to comply with Ontario’s Employment Standards Act, namely the AI provisions in Bill 149, in the near future. As of some future date to be named by proclamation, employers making publicly advertised job postings will be required to disclose within the job posting if they use AI to screen, assess or select applicants for publicly advertised job posts. These employers will also have to retain copies of every publicly advertised job posting (as well as the associated application form) for three years after the post is taken down. 

In fact, I recently wrote an article on this topic. I wrote about how skeletal the AI provisions were in this bill. And the AI-related definitions were nowhere to be found. I compared the requirements in Ontario’s Bill 149 to those in New York City’s (NYC’s) hiring law involving using AI and automated decision-making tools. It was striking that NYC required employers to conduct a bias audit before using the tool; post a summary of the results of the bias audit on their websites; notify job candidates and existing employees that the tool would be used to assess them; include instructions for requesting accommodations; and post on their websites about the type and source of data that was used for the tool as well as their data retention policies. There were detailed definitions and fines for noncompliance too. 

Needless to say, this will be a challenge for provincially regulated employers in Ontario, and it is highly recommended that employers prepare now for these employment law changes. That said, it is understandable that employers may struggle with how to comply with such ambiguous provisions. 

Additional issues could arise, namely privacy and human rights issues. Let us take the example of the video interview where sentiment analysis is conducted. This is troubling from a privacy perspective—job applicants may not be comfortable participating in video interviews where their facial expressions and gestures are closely scrutinized with intrusive software that enables AI tools to analyze their sentiments. Moreover, employees who are up for a promotion may not appreciate video analytics of their video interview performance being retained for an unknown period of time, and accessible to an unknown number of actors in the workplace. Because job applicants are in a vulnerable position, they may not feel like they can object to the use of these AI hiring tools.  

In addition to privacy concerns, human rights issues could surface. The video interview could reveal various aspects of a person that may fall under any of the prohibited grounds of discrimination. For instance, under the Ontario Human Rights Code, section 5 of the Code prohibits discrimination in employment on the grounds of race, ancestry, place of origin, colour, ethnic origin, citizenship, creed, sex, sexual orientation, gender identity, gender expression, age, record of offences, marital status, family status or disability. 

It may be possible for an AI tool to be biased (unintentionally, but biased nonetheless), where it favours the younger candidates, gives them higher interview scores, and ultimately inadvertently discriminates on the ground of age. Since it may not be possible to detect these biased decisions immediately, it may be that some job applicants simply miss out on an employment opportunity due to an ageist AI tool.  

It will be interesting to see whether other jurisdictions come up with more extensive provisions to address the use of AI in recruitment and hiring. In Ontario, it is questionable whether we will see additional detail to help employers comply with the requirements.