JD change/update for AI

[Mindy McManus]

#108384

Forum: Private

Hello,

I am wondering if there is an obligation to change JDs now that our software development team is using AI at work. Should we worry about time and training for the new AI skillsets for our developers job in case we need to make terminations in the future?

Thanks!

[Haley O’Halloran]

#108387

There is generally no automatic legal obligation to amend existing job descriptions simply because developers have begun using AI tools. That said, from an HR and future performance-management perspective, I would recommend reviewing the JDs if AI has become a meaningful part of how the development work is expected to be performed. The key distinction is between AI being merely another tool used to perform existing duties versus the employer introducing genuinely new competencies or substantially changing the core role. Ontario recognizes that significant unilateral changes to core duties and responsibilities can, in some circumstances, raise constructive-dismissal issues, so major changes should be handled more carefully.

The bigger issue for future terminations is expectation-setting and training rather than the wording of the JD alone. If, for example, you later conclude that a developer is underperforming because they cannot effectively use AI-assisted coding, prompting, code-review or testing tools, it will be much easier to support that decision if you can show that the competency was clearly communicated, employees were given reasonable access to the technology and training, and they had a reasonable opportunity to demonstrate the required standard. I would therefore document the new expectations through updated JDs, competency profiles, performance goals or development plans where AI proficiency has become material to the role, and provide reasonable working time for employees to learn the tools rather than expecting them to acquire substantially new skills on their own time.

I would also avoid framing a future termination as simply “you don’t have AI skills.” Instead, identify the actual job outcomes involved—for example, ability to use approved development tools efficiently, validate AI-generated code, protect confidential information, identify errors or security vulnerabilities, and meet established quality/productivity standards. That creates a much stronger performance-management record and also avoids tying employment decisions to a particular technology that may change quickly. One additional Ontario point: as of January 1, 2026, certain employers with 25 or more employees have AI-related disclosure obligations for public job postings where AI is used to screen, assess or select applicants, but those provisions do not themselves require employers to update existing employees’ job descriptions merely because employees use AI in their work.

-HRInsider

[Author]

Posts