Can Publishers Use Metadata to Regain the Public’s Trust in Visual Journalism?

The News Provenance Project has been exploring how news organizations might contribute to the fight against misinformation by adding context.

Illustration by Jinhwa Jang

“The thing about photos is that they carry with them a lot of credibility. Someone took a picture of something! The fact that you can have a genuine photograph, but the way it’s being used is bogus — that’s interesting.”

These are the words of Roger (not his real name), an avid newspaper reader and TV news watcher who also keeps up with the news through Twitter. Roger was generally aware of the prevalence of misinformation on the internet, but he was surprised to learn that it affects the photos he sees — and that he could easily believe images that are accompanied by false information.

Roger’s realization came during a conversation with Emily Saltz, a user experience researcher and designer for The News Provenance Project. Emily had been showing him an example of a social media feed where photos had been posted with inaccurate information about the events shown. His reaction was far from unusual in the more than 30 interviews Emily did as part of our research into how American news consumers from a variety of backgrounds and political leanings make sense of the images they see online.

That it is easy for people to be swayed by fake content should come as no surprise. The arenas in which news and information appear these days are messy ones, with social media platforms making no visible distinction between legitimate posts and deliberately misleading ones. Almost every major news event brings a new set of examples: as Buzzfeed News reported in the last few weeks, multiple instances of misinformation emerged both after a United States airstrike killed a top Iranian general and after Iran retaliated with airstrikes on American military bases in Iraq.

And critically, misinformation continues to disrupt elections around the world, as it did in the run-up to Britain’s general election last December, and threatens to interfere with the United States presidential race this year.

What might clear up this polluted environment? And how can news outlets help audiences better understand the legitimacy of what they publish? While there is no one-size-fits-all solution, the urgency for creating effective interventions continues to grow.

We launched The News Provenance Project in mid-2019 to address the misinformation crisis through a product and reporting lens. Our goal was to contribute to the work of a growing number of organizations and projects addressing this issue. The initial idea was that publishers could contribute to creating a healthier information ecosystem by surfacing information they already have about the work that they publish. Our main hypothesis was that adding context to news photos — in the form of metadata, with information that is often contained in a caption, but gets stripped out as a photo travels beyond a news outlet’s own sites and apps — might help people make better decisions about the credibility of the images they see on social platforms and elsewhere around the internet.

Today we are publishing the results of the research we conducted, and a few examples of best practices for designs that help people better understand what they see. A key component in that research is a proof of concept, which we’re also making available today. The News Provenance Project team, working out of the New York Times R&D Lab, built it as an example of how our hypothesis might be realized, and we developed it in response to several rounds of user testing. We also explored the feasibility of publishing the data associated with each photo, from which that context is derived, on a blockchain (more on that below).

As we develop plans to expand our work in 2020, we hope that sharing this research might also inspire others to build upon it. But first, a bit of background:

Why we used blockchain

In addition to serving as a visual example of how provenance information could be added to the presentation of photos, the proof of concept gave us an opportunity to experiment with blockchain as an emerging technology, one that lends itself well to the features we wanted.

As a database, a blockchain can provide network members increased confidence in the reliability of its data, due to its enhanced security in the way it stores that data, which makes the records published to it more or less immutable. It also ensures transparency of all “transactions,” or updates to the records stored in the system, so that any major changes made to a record would be recorded and visible. In addition, it offers shared ownership of a database among a number of entities — in this case, publishers.

Here’s a hypothetical scenario on how those properties could become useful: Let’s say five news publishers agreed to participate in a network and publish some or all of their photos through a News Provenance database. All five, and potentially also the public, would be able to see the metadata of the others, as well as any changes made to those photos after their publication. Those changes would not be alterable by anyone.

The News Provenance team — mainly Emily Saltz and Pooja Reddy, a product manager at The New York Times — worked with technologists and designers from IBM’s Garage to build the proof of concept on IBM’s blockchain platform, which uses Hyperledger Fabric, an open-source blockchain technology. We set up the basics of a data structure and a network and refined the designs, then published a number of photos and their metadata to the database. That work is behind-the-scenes of the proof of concept.

[Go behind-the-scenes into the research, design and prototyping behind The News Provenance Project’s proof of concept.]

From the user’s perspective, there is no discernable sign of how the data is stored. In our original hypothesis, we guessed that those properties, if represented clearly to users, could encourage their trust in both the legitimacy of the photos. However, we learned from the people we interviewed that words like “immutability” and “encrypted database” were not important factors in establishing trust and confidence.

If we were to build out such a network in the real world, we’d need to explore the feasibility on a few fronts. The operating speeds may not be as fast as we would need them to be and costs might be high, which are two barriers to news organizations, especially smaller ones, for whom slow speed and high expense might be prohibitive. And a system of governance would need to be put in place so that all the organizations on the network agreed to work with it in the same way, both in the data they would provide with each file and how they would represent it on their own sites.

Helping Audiences by Showing Our Work

We don’t expect that a layer of additional information alone will be enough to change the minds of people who distrust a given news brand, or that it will help them completely avoid being swayed by false context and manipulated media. But it could help.

Along with the reactions garnered in our own research, there are other signals that transparency helps readers. According to Trusting News, a project that has done research and training for newsrooms on what helps readers trust news outlets, labelling and explaining the journalistic process are two ways that can help engender audience trust.

Critical information that typically goes into a photo caption, such as time, date, location and the accurate identification of people and events shown doesn’t travel with a photo when it’s posted to social media, where it can be reposted with egregious inaccuracies. News organizations have this information. With a bit of work and some thoughtful designs, we could use it to help platforms and, more importantly, news consumers avoid inaccurate uses.

Innovating in Reverse

Our proof of concept only focused on the point at which a photo is published. To ensure that the verification work done by reporters, photojournalists and their editors is not lost once a photo gets shared on social media, we need innovation at every stage of the process, from photographing (or capture) through publishing, through distribution. However, to put it simply, this is easier said than done.

To give a sense of what it would take to realize the full vision of surfacing provenance information on news photography, we’d need changes at every step of the process, from the time a photo is taken, to every instance of its publication and display. In brief, an ideal set of changes would be for:

  • Camera makers to help photographers ensure time, date and location settings in cameras are exact.
  • Every news publisher to modify their management processes for photo metadata so that they adhere to a common set of standards, such as those maintained by the International Photo Telecommunications Council (IPTC).
  • All platforms such as Google, Facebook, Twitter and Apple, as well as chat apps like WhatsApp and Signal, to ensure the consistent display of this information.

Within this big and complex endeavor, a proof of concept is a small drop in a very big, complicated bucket — but before taking bigger steps, our goal was to explore its feasibility, including how it might resonate with real people who encounter misinformation online every day. We’ve seen that it could have an impact with people of varying ages, from different geographic locations and party lines. We are eager to bring it to life with further experimentation in the future.


The News Provenance Project is grateful to a number of people who contributed to our work in 2019:

  • IBM’s Garage team were excellent collaborators who worked with us on the design and implementation of our ideas.
  • Bernat Ivancsics, a PhD candidate in communications at Columbia University School of Journalism, joined our small team as summer fellow through the Brown Institute of Media Innovation. Having done earlier work on blockchain in journalism, Bernat came with a depth of knowledge and a breadth of thinking about the topic, and helped us grasp the wider context and the workflow considerations.
  • Bhaskar Ghosh, a dual-degree masters candidate in journalism and computer science, also at Columbia, was also a summer fellow through the Brown Institute. His work on a Chrome extension that could surface metadata as an overlay on photos on twitter provided an initial visualization at a time before we had any designs, and demonstrated another potential route to the display of provenance. He also explored the technical feasibility of identifying publisher photos and researched perceptual hashing and computer vision algorithms.
  • Niko Koppel, a visual journalist and photo editor who works on the Times R&D team’s 5G projects, provided photos of scenes he had shot while on assignment, as well as critical insight into the photo assignment and editing process.
  • Representatives from several news organizations, including the AP, Wall Street Journal, Hearst and others, contributed to a design thinking workshop in the spring out of which came the initial ideas and problem statements.

Sasha Koren is an independent editorial consultant who has served as project lead on The News Provenance Project. Previously, she was the editor and co-lead of the Guardian Mobile Innovation Lab and held a number of digitally focused roles in The Times newsroom.

Can Publishers Use Metadata to Regain the Public’s Trust in Visual Journalism? was originally published in NYT Open on Medium, where people are continuing the conversation by highlighting and responding to this story.

Source: New York Times

Leave a Reply

Your email address will not be published.