Article content

The resumé used to be a document. It is becoming a performance.

Sign In or Create an Account

or View more offersArticle content

Artificial intelligence can now take an ordinary employment history and turn it into an impeccably structured account of accomplishments, leadership and strategic thinking. It can tailor the language to a particular employer, optimize it for candidate-tracking software and prepare the applicant for almost any predictable interview question.

Article contentWe apologize, but this video has failed to load.Try refreshing your browser, or
tap here to see other videos from our team.Article content

Story continues below

This advertisement has not loaded yet, but your article continues below.

Article content

The employer then feeds that resumé into its own software to decide whether the prospective employee is worth meeting. We have created a peculiar system in which one computer helps an applicant persuade another computer that the candidate is exceptional.

Article contentArticle content

And employers should be worried. Not simply because AI can make mistakes, but because employers may conflate fluency with competence.

Article content

Ontario has responded with a new rule that came into force Jan. 1. Companies that have 25 or more employees and use AI to screen, assess or select applicants for publicly advertised positions must disclose that fact in the job posting.

Article content

That is sensible. An applicant should know when technology is being used to decide whether they get through the door.

Article content

But what does the law not address?

Article content

It does not require the employer to explain how the system reaches its conclusions. It does not require disclosure of the criteria being applied. It does not require proof that the system produces accurate results.

Article content

And the applicant does not have to announce that their beautifully written resumé was produced by ChatGPT.

Article content

Story continues below

This advertisement has not loaded yet, but your article continues below.

Article content

The law regulates one side of an increasingly technological contest while saying little about the other.

Article contentRead More

  1. Howard Levitt: New workplace monitoring software has employees up in arms. They’re concerned about the wrong thing
  2. When a disability such as addiction won’t shield employees from being terminated for cause
  3. Story continues belowThis advertisement has not loaded yet, but your article continues below.

Article content

That is where things get interesting. Consider what AI recruitment systems are often being asked to do: find candidates who are similar to the people the company has historically hired. That sounds reasonable, until you ask a more uncomfortable question:

Article content

What if the companies historical hiring decisions were not producing the employees the company now needs?

Article content

An algorithm trained on yesterday’s hiring decisions can turn yesterday’s preferences into tomorrow’s screening criteria. A company that historically hired mostly younger employees may, without intending to, build a system that regards the career histories of older applicants as less attractive. That is illegal.

Article content

A company whose successful employees followed conventional career paths may disadvantage candidates who took career breaks, changed industries or acquired experience outside the traditional route. That does not automatically establish discrimination. But it creates a legal question that an employer cannot answer by saying the computer was objective.