The streets of downtown Montreal are cluttered with protestors chanting, “Fuck the police!” Traces of the sun filter out from behind department store buildings as anti-capitalists rally for International Workers’ Day. Spectators capture footage of police spraying a thick cloud of tear gas into the crowd, which sends people running. Marie-Espérance Cerda interviews protestors and documents the unfolding events, but instead of using a standard video camera, she’s using six GoPros on a rig to produce virtual reality (VR) journalism.

Later, electronic headgear that creates a three-dimensional, interactive environment will immerse viewers in the same scene Cerda witnessed. This experience—the feeling of being in a place and the heightened sense of emotion that goes with it—isn’t possible through traditional journalism. Words on a page or a video on a screen creates distance between the reader and the story, an empathy divide that VR shrinks.

While the concept has been around since the Second World War, long before computer scientist Jaron Lanier coined the term “virtual reality” in the 1980s, its most common commercial application so far has been video games. In the past year, though, journalists have explored VR’s powerful storytelling possibilities, but they must navigate the tricky ethics that invariably come with new technology.

A VR headset creates a 360-degree field of vision that moves with the user, allowing her to explore virtual surroundings and become part of the story. Lenses focus and reshape the display to make a three-dimensional stereoscopic image similar to one in a View-Master toy. Most high-tech headsets take measurements of the user’s skull to record motion, giving the user control. In 2014, Google released Google Cardboard, a build-it-yourself device that makes VR accessible to anyone with a smartphone. A small magnet works with the phone’s magnetometer (which controls the compass) to create movement.

Cerda’s Montreal experiment began as a major research project for her master’s degree in media production at Ryerson University. The 10-minute video starts outside of a downtown Burger King. Straight ahead, people wave Quebec flags and hold picket signs high. Look up and you’ll see the remnants of daylight reflected in a partially blue sky. If you turn around, there’s a white bus parked in the middle of an intersection. Police armed with riot shields file out one by one. Then they start spraying tear gas.

The coolness of VR can overshadow ethical concerns. There’s more control, but the viewer is confined to the passenger seat. “You’re existing in a universe of possibility that’s been defined by the person who’s made the news item,” says Gene Allen, a journalism professor at Ryerson and the supervisor for Cerda’s project. “They’ve decided what to shoot, and they’ve decided how to put it together.” While “inside” the protest video, viewers can pick where to look and whom to listen to—an illusion of choice. But there’s limited perspective on what’s happening outside the frame. There’s a similar selection process in all forms of journalism, notes Allen: reporters include what’s interesting and toss the rest.

In November 2015, The Globe and Mail launched a roughly three-month VR trial. Three employees spend their days inside an incubation lab on the main floor of the paper’s building. A lot of the current focus is on the technological aspect, says Matt Frehner, senior editor of mobile and interactive news, adding that the VR team is still in the “how does this work” phase. The goal is to create an immersive experience that’s as different from regular video as IMAX is from a regular movie. Meanwhile, Canadian Press plans to explore the technology’s potential within the next year.

Still, Canada is a few steps behind American outlets. ABC and The Wall Street Journal have created VR content. And last November, Associated Press announced plans to produce a series of downloadable stories, which will be released by March.

On Sunday November 8, 2015, The New York Times arrived with Google Cardboard, allowing subscribers to watch an 11-minute video called The Displaced. It followed three child refugees, including 9-year-old Chuol. When his village in South Sudan was attacked, he fled to the swamp with his grandmother; his father and grandfather were burned alive, and he was separated from his mother. He stands at the front of a hollowed-out wooden boat, paddling through a narrow stream surrounded by thick blades of grass and lily pads. The sun reflects off the water, which may conceal crocodiles—an ever-present threat in the swamp. “I know that if I am eaten by a crocodile, it may be a slow death,” the boy says in the video, “but it is better than being killed by the fighters.”

Stories told through VR are usually emotional ones, and the danger is some will go too far. Would people want to experience the terrorist attacks in Paris? The earthquakes in Nepal? Empathy is a powerful tool, when used correctly, and VR breaks down familiar barriers that stand in the way of complete understanding. In the Times project, instead of trying to imagine what living conditions are like in South Sudan, VR lets people temporarily experience it for themselves. Feelings are enhanced and perceptions are amplified, but that can push people into dark corners.

After the paper launched the VR project, Michael Oreskes, news chief at National Public Radio and a former Times editor, was among the people who voiced concern. “Our stories can’t be virtually true,” he wrote. “They must be fully real.” While some projects (including Cerda’s video and the Times’s refugee film) are made from real-time footage, others use computer-simulated images based on maps and photographs. But can embellished stories be honest stories? Allen believes they can, so long as reporters clearly indicate what they’re doing. Feature writers reconstruct scenes all the time, he says, and television programs, including CBC’s the fifth estate, often use simulated footage. The difference with VR is that it’s harder to draw the line between what’s real and what’s recreated. It’s up to the journalist and the editor to produce content that serves as a genuine representation of a story.  

Late last September, Cerda presented her VR project at Ryerson to a small group of people huddled around a boardroom table. A woman strapped on the cardboard headset and became immersed in cluttered Montreal streets as people chanted and police filed out of a white bus. She spun around in a black office chair and said, “Incredible.”