The Canadian Association of Journalists (CAJ) is deeply concerned by the recent publication of an AI-generated image on the front page of the Winnipeg Sun.
On May 13, the Winnipeg Sun published an AI-generated image combining scenes of a homeless person receiving a disposable cup from a person outside of a church and a crowd of sign-carrying protesters.
A similar image was published online on the same day. The image carries the following cutline: “A symbolic illustration depicts tensions surrounding homelessness services, charitable giving and organizational change amid financial pressures facing Winnipeg’s Siloam Mission. (AI-generated image).”
“In a world where deepfakes, misinformation and disinformation are running rampant across social media and the news media landscape, using a misleading AI-generated image sets a dangerous precedent,” said CAJ president Brent Jolly.
The CAJ’s recently-released AI Ethics Guardrails discussion paper states: “AI should be treated as a tool that complements journalistic judgement – never as a replacement for it.”
The problems with the Sun’s use of the AI-generated image are twofold.
First, the image is clearly misleading. It is designed to emotively agree with the writer’s view that a change in the mission’s focus to more actively try to embrace reconciliation with Indigenous peoples has led to a collapse in donations. While that is the writer’s opinion, the column does not cite any examples of the public protest that is referenced in the AI-generated image.
Second, the publication’s use of artificial intelligence to generate new content is nowhere near transparent enough. While the CAJ notes that the cutline does indicate that the image was generated with artificial intelligence, its use, particularly on an editorial page, necessitates far more prominent disclosure.
The CAJ reached out to the Winnipeg Sun to ask for a response to the ethical concerns.
“We agree that the use of the image in question was inappropriate and did not meet the standards we expect of our newsroom,” Kevin Klein, CEO of The Klein Group, which owns the Winnipeg Sun, told the CAJ.
“Our team has been clearly informed and is now reminded of our policies regarding the use of AI-generated content; however, in this instance, a staff member made an error in judgment. The matter has been addressed internally,” said Klein.
As journalists, editors, reporters, photographers, and newsroom leaders, we have an abiding duty to inform our readers with the facts of a particular story, including the use of images and videos. The Winnipeg Sun’s use of a misleading, artificially generated image is a clear violation of that duty.
“Artificial intelligence tools must be used with caution, care and always with clear editorial policies in place,” Jolly said. “The public’s trust in our work depends on it.”
The CAJ urges all reporters and editors to abide by the ethics guidelines on AI image use, which can be found on our website: AI Ethics Guardrails. Further unethical AI image use can be reported to CAJ President Brent Jolly at b.jolly@caj.ca.
The Canadian Association of Journalists is the country’s largest professional organization that serves to advance the interests of journalists from coast to coast to coast. The CAJ’s primary roles are public-interest advocacy work and professional development for its members.
For further information: Brent Jolly, president, Canadian Association of Journalists, brent@caj.ca
Donate
CAJ Strategic Plan 2024-2027
