Canadian public servants issued guidelines for using generative AI
The Canadian federal government has rolled out a set of guidelines to help officials use generative artificial intelligence (AI) tools responsibly at work.
The preliminary guidelines define generative AI as an information technology “that would ordinarily require biological brainpower to accomplish, such as making sense of spoken language, learning behaviours, or solving problems”.
They were announced to mark “the initial stages” of the Canadian public service’s “[recognition of] the importance of AI”, according to Anita Anand, president of the Treasury Board of Canada.
Designed to complement government’s existing directive to departments on AI, the new rules are expected to undergo updates over time. Their core aim, is to tackle instances where the use of AI constitutes malpractice, resulting in a cybersecurity threat, misinformation, or where generative AI tools like ChatGPT and Google Bard produce discriminatory outcomes.
Anand said she welcomed the introduction of the guidelines, describing herself as “a racialised woman” who understood the risk of discrimination firsthand.
“I will say that the purpose of these guidelines is to ensure responsible use of generative AI and we will be monitoring to ensure that bias does not creep in if employees do go down the road to use generative AI,” Anand said.
There are currently no penalties for violating the guidelines, though Anand stressed that the new rules were based on existing pieces of legislation, such as the Privacy Act. A violation could subsequently lead to a penalty, depending on the nature of the transgression.
She said employees would still have the same legal obligations as before “regardless of these guidelines”, and that the new rules came “on top of those existing obligations”.
Read more: UK to host global AI safety summit at iconic WWII codebreaking HQ
Government to improve transparency of AI use
The publication of the guidelines comes after Global Government Forum research revealed that more than 10% of Canadian public servants say they have used artificial intelligence tools such as ChatGPT in their work.
The guidelines demand that government departments always make it known to citizens if a service uses AI to interface with a user. This applies to any service that uses generative AI to respond to a citizen’s query, create a document or make a decision.
Departments will also be required to “identify content that has been produced using generative AI, notify users that they are interacting with an AI tool, document decisions and be able to provide explanations if tools are used to support decision-making”.
“These guidelines…will make sure that employees are aware of not using private or secret information, making sure that content is factual, making sure that we are transparent about its use, and making sure that we’re complying with laws and policies as well.”
Read more: One in ten Canadian public servants already using AI for work purposes
Call for government to be clear about ‘no-go zones’ for AI
Michael Wernick, former clerk of the Privy Council of Canada and secretary to the Cabinet, told Global Government Forum that while he expected discussions around AI in government to skew towards externally facing services, he saw greater potential for AI to transform internal processes related to Human Resources and procurement.
“It now seems possible to load an AI engine with every collective bargaining agreement, every job description and every selection profile in the whole public service going back a decade and train it to write new job descriptions, draft competition posters, sort applications into piles, mark simple written exams and write letters of offer and rejection,” he said.
“[This] could take away the excuses for tardy staffing.”
Jennifer Carr, president of the Professional Institute of the Public Service of Canada, took issue with the speed with the which the guidelines were introduced, saying her union (which includes government computer experts) had not been properly consulted. She added that the guidelines did not contain sufficiently clear delineations of good and bad practice.
“It only proposes that the government ‘be careful’ on how they use AI. “‘Be careful’ is a very subjective term,” she said.
“What we’re really looking for is that there are strict regulations or guidelines, where there are go and no-go zones.”
Carr said finally that public servants and citizens alike would need further reassurance that AI did not put jobs and services at risk by outsourcing decisions to computers.
“If somebody is cut off of a benefit because there’s no human interface, that’s where we get concerned” she said, adding that government had a duty to “upskill” employees to take effective roles in new AI systems.
In response to these concerns over the future of federal jobs, Anand said: “This is not about replacing employees at all.”
“This use of generative AI is as a tool to further the work of existing and future employees.”
Read more: Canadian government spending cuts being ‘rushed’, warns union
About Jack Aldane
Jack is a British journalist, cartoonist and podcaster. He graduated from Heythrop College London in 2009 with a BA in philosophy, before living and working in China for three years as a freelance reporter. After training in financial journalism at City University from 2013 to 2014, Jack worked at Bloomberg and Thomson Reuters before moving into editing magazines on global trade and development finance. Shortly after editing opinion writing for UnHerd, he joined the independent think tank ResPublica, where he led a media campaign to change the health and safety requirements around asbestos in UK public buildings. As host and producer of The Booking Club podcast – a conversation series featuring prominent authors and commentators at their favourite restaurants – Jack continues to engage today’s most distinguished thinkers on the biggest problems pertaining to ideology and power in the 21st century. He joined Global Government Forum as its Senior Staff Writer and Community Co-ordinator in 2021.
Related Posts
