The Government of Canada is working to develop a framework to address discrimination in recruitment processes within the government sector, specifically addressing issues of privacy and transparency.
This comes as several departments have begun incorporating AI into their hiring processes, despite numerous complaints being raised about the discriminatory nature of AI tools.
respect for privacy
Finance Commission President Anita Anand acknowledged the transformative power of AI in an interview that covered a wide range of issues. She also acknowledged concerns about its implementation and called on the Government of Canada to develop a set of guidelines to limit discrimination.
“There is no question that individual privacy must always be respected in accordance with privacy laws and employment practices must be non-discriminatory and have a sense of equality built in,” Anand said. he told CBC.
“Certainly, as a racialized woman, I feel this very deeply. There is a need to ensure that the use of AI in the workplace complies with current law and stands the moral test of not being discriminatory. there is.”
Also read: Microsoft AI provides misleading answers to election questions
Skill assessment using AI
In response to a question from NDP MP Matthew Green, the Department of Defense indicated it was using an AI-driven skills assessment tool. Nokkuri.
The tool was used in hundreds of pre-recorded job interviews as part of an effort to “reduce bias and promote fairness in the hiring process.”
“Knockri utilizes machine learning to analyze the audio-to-text content of candidates' responses and compares them to determined performance metrics,” the agency wrote.
Nokkuri is a Toronto-based company that has developed an AI-based video recruitment tool to limit bias and enhance the hiring process.
Is the government spying on its employees?
Anand's interview with CBC News came just a month after the investigation by. radio canadarevealed a privacy violation of personal data.
It has emerged that government departments have acquired technology that can be used to extract data from mobile devices such as cell phones and computers given to employees.
Mr. Anand denied allegations that the government was using spyware on its employees.
“This is something that is only used when there is an investigation or if we think something is going wrong,” she said, adding, “It's not something that would be used on a daily basis.”
AI discrimination is a long-standing problem and a challenge not only in Canada but globally. In September, Anand introduced guidelines for civil servants wishing to use generative AI tools like ChatGPT, saying the government would be mindful of potential challenges such as discrimination and bias.
But Global Affairs still apologized for posting AI-generated content representing Inuit women on social media, according to CBC.
Global Affairs should have used tools created by the Treasury Department “such as AI guidelines and the Algorithmic Impact Assessment (AIA) process” to guide employees and departments.
“This is a perfect situation to use the algorithmic impact assessment tools I described,” she said.
“The AIA tool is a series of questions that need to be asked when AI is being used in the workplace. These reflect legal, policy and ethical considerations. ” Anand added.
She also expressed a desire to fill gaps in her country's AI regulatory framework.
“There's no question that AI is a transformative technology. Another thing I'm looking at is really long-term, and that's about regulation and more broadly how we Because we need to ask how we think.”
“For example, with the advent of autonomous machines such as self-driving cars, we have to ask ourselves whether the regulations we are putting in place are appropriate and applicable in the long term,” Anand said.
Humans will remain relevant
While the government is considering AI-powered tools within the ministry, Anand said Canadians should always have the option of talking to a live human being. Therefore, despite adopting this technology, civil servants should also not be negatively affected by the implementation of AI technology in their departments.
This comes in the wake of fears and fears that AI technology will take away jobs and put millions out of work.
Through Bill C-27, which is currently before committee, the Government of Canada wants to adopt a framework for the use of AI in federally regulated businesses. However, Anand did not say when the law would start applying to government departments.
But in 2024, Anand will be tasked with revamping the aging technology that governments use to deliver benefits such as the country's pension system and employment insurance, Old Age Security (OAS). .
In line with this, the government transferred 600,000 foreign OAS recipients to the new platform for the first time in June. The second phase of the exercise is scheduled to begin next year.
“Research shows that there is a direct correlation between the trust that citizens have in their government and the services they receive. We are hopeful that we will see visible progress,” Anand said.