The slow and steady rise of AI for news

How can publications boost their content with AI and automated tools?

Emilie Kodjo
Apr 27, 2017 · 9 min read

Artificial intelligence and automation have recently crossed over into mainstream territory, carving an ever-growing space for themselves into newsrooms, allowing journalists to produce articles with highly shareable content and, putting the newsroom and its content at the forefront of social media platforms, in the quest for answers to the questions: How to bridge the gap between the print audience, the website audience and the highly-coveted eyes of Gen-Z and millennials on social media? We talked to Zohar Dayan, co-founder of Wibbitz to try and understand how AI and automation are helping publishers.

GEN: How have AI and automation evolved in newsrooms recently?

Wibbitz: We’ve seen in the US that in the past year or so there was a strong shift in the industry, a much more openness to more innovative technologies in the newsroom, especially with AI, it is something that was a taboo before with news companies that were naturally very traditional. Especially editors were very hesitant when it comes to artificial intelligence creating content. We have been seeing a huge shift, an evolution in the market with much more acceptance towards AI as a recommendation tool, AI that writes articles from data, infographics generated automatically using data feeds, and in our case, automatically produced video content.

All of this is a testament to business models, newsrooms and media companies’ that strive to be more efficient from a business point of view, and also in order to keep up with the high user demand for content and the diversified platforms that newsrooms are catering for.

We’ve been operating in several European countries, in addition to France, we work with Germany, Spain, Italy. We are opening an office in Paris as we really sensed that the market in France is becoming very innovative fast. We managed to start working with very significant companies, that you would not necessarily expect to be so innovative. For instance TF1 which is by all means a traditional TV broadcaster, that has recently been focusing on the digital space, evolving into the next stage, shaping up their future around digital content. It’s interesting, we have been seeing a lot of interest and traction from big companies, in addition to TF1, Le Parisien, Le Figaro. All were very interested in the technology and implemented it very quickly. It is very similar to the trend we saw rise in the US about a year and a half ago.

How do you approach implementing AI tools within the newsroom?

When we first started we were very focused on the automation aspect –which we still are – but there were a lot less editing tools and flexibility within the creation process so over time we worked very closely with all of our partners so we got a lot of time to get feedback from them, show them ideas of what we were thinking of adding to our roadmap, as a sort of a discovery process, or whenever we were making changes to the product. Through these exchanges we started to see how a publishing company perceived seeing automation: the perception from 2 years ago to now is very different, it is amazing to us.

Seeing that a publisher could be more open-minded to tools like ours. We have been able to build our product in a way that works well with what reporters are looking to get out of an automation tool: We see it as a combination of human and machine.

“We have been seeing a huge shift, an evolution in the market with much more acceptance towards AI.”

How has the news industry changed over the past couple of years in its approach of automation and AI?

Two years ago, when we first introduced this idea to a publisher like Forbes, it wasn’t that they didn’t like the tool it was that they weren’t really open to giving it a chance, from what we saw in the situation. They did entertain the thought that it was perhaps useful, without fully believe or understand it themselves. And didn’t really try to integrate it into their workflow. We were met with strong resistance. There was often an element of fear when talking about automation and AI, especially with content creation. It happened with many industries though, whether it was for Uber with self-driving cars or product like ours.

Now, automation is better understood and far less threatening, as a process going through a tool which is no way a replacement for a person. There is a lot more acceptance now, which has enabled us to work with most of the publishing companies in the US and many in France, we are expanding in Europe and Australia.

Getting over the initial hump of fear and resistance was crucial, and as soon as we got the opportunity to show how the tool can be beneficial without reducing the quality of what newsrooms and publishers are doing, publishers got on board more easily.

It is very much like using the comparison of power steering: automation will allow newsrooms to do their job in a faster way. It has been a big change. And perceptions have changed over the last couple of years, how publishers and journalists are now viewing automation and AI: how receptive they are to it, they see it as useful to them.

We see more opportunities in the future for more complete automation, having the entire creation process automated, it still won’t be a replacement for people and human interaction, but as the demand for videos shows no sign of slowing down, incorporating automation as a tool means to ease and speed up processes.

Was there something that surprised you in the use that publishers, editors or journalists are making of Wibbitz? Was there something that you didn’t quite anticipate?

One of the major uses we did not account for when we rolled out our product was the use of video on social media, which is not anything new, but we saw how a publisher could use our platform to adjust videos for every platform: automatically produce a square video or a vertical video. We have a set template to create automatically a 10-second vertical video, that looks very similar to what you see on SnapChat Discover or on Instagram Stories. This was a direct response to what we were seeing on the market. Publishers having a difficult time creating videos for all these different platforms in the right format for each one. That is extremely time consuming. New AI and automated products to help support publishers do just that.

“There was often an element of fear when talking about automation and AI, especially with content creation.”

Do you think it helped some publishers and newsrooms to embrace Snap, Instagram Stories or the social media features that weren’t necessarily obvious to them?

Definitely. We worked with TMZ and a few other publishers when we first rolled out the ‘Snippet’ vertical video. They started using it for Instagram, right when Instagram Stories started. It is something we suggested to them, as they weren’t creating content for it before.

This is a phenomenon we see a lot more of for mid-size publishers as they adopt automation, as they might not have thought to create content for Instagram and Snapchat. It has become a good opportunity for them to start publishing on these platforms and reach new audiences. An increasing number of publications follow this route, as they fully complete their transformation to digital.

What should newsroom do if they want to start with AI today?

It is sometimes difficult, but it is becoming less just because of the market, the need for video, the rise of the importance of video, for social media and general reporting, is everywhere. Publishers now fully understand that they need to do this, they need their content to be as flexible as possible and published across an array of platforms: their own website and their social channels, to engage their audience and convert new ones.

A lot of publications do not have an in-house video team understanding how to create these very short, ‘snackable’ type of videos. Although they might do long-form documentaries, or might be broadcasting companies, they do not necessarily understand the types of video content you can create through automation.

The work is getting more and more cut out for us to educate publishers on automation and AI because these topics are discussed, adopted and are becoming more and more mainstream, discussed and adopted.

What would be your advice for a startup looking to introduce a tool in the news industry?

Education on your product is crucial, but also a sound understanding of not only the market but the problematics publishers, editors and journalists are faced with every day. Also turning your prospects or clients into partners: with a true partnership going both ways, going beyond the regular client-supplier relationship, to be as close as possible to the demand, to understand rising issues and topics a newsroom meets so you can adjust your roadmap to meet the needs of the newsroom. This is the best way to carve a space for a startup. It is a hard and a long process, but eventually things start to shift.

How do you foresee AI shape the future of news?

AI will continue to develop and make breakthroughs in more and more industries. Specifically in news, AI will become an essential part of making a more productive and efficient newsroom, by speeding up processes. AI will empower journalists and editors to make better decisions and automate many of the time-consuming and labor intensive tasks that are not directly related to the content. AI and automation will enable storytellers to focus more on the craft of journalism and less on the mechanical work, leading to a more visual and rich news experience. Humans and machines working together will create a more streamlined workflow that answers the evolving content consumption patterns that we’re seeing today and will continue to see in the future.

What elements need to evolve for newsrooms to fully embrace AI and other new technologies?

  • Editors and journalists should embrace these technologies and not fear them. Yes, AI will change the way they work but in the long-run it will empower them to reach their users in more effective ways
  • Newsrooms must create in-house innovation teams that are constantly searching for innovative technologies that they can quickly test and validate. They need to be much more connected to the startup community, plan hackathons and always be on the hunt for new trends.
  • Changing habits! Many newsrooms have become too used to their old way of thinking. For them to survive they need to mix things up and make drastic changes, sometimes even painful ones. The landscape is changing fast and so should they if they want to keep up.

About Wibbitz

Wibbitz is a text-to-video creation platform built for publishers. Its advanced text-to-video technology can automatically produce premium branded videos using text content in seconds. Wibbitz just opened a new office in Paris.

Damian Radcliffe—University of Oregon

“Bots and automation are increasingly becoming a part of how journalism is produced and content is being consumed.” (, 5 November 2016)

John Micklethwait—Bloomberg News

“Done properly, automated journalism has the potential to make all our jobs more interesting.” (The Irish Times, 23 March 2017)

Lou Ferrara—AP

“Automation was never about replacing jobs. It has always been about how we can best use the resources we have in a rapidly changing landscape and how we harness technology to run the best journalism company in the world.” (The Huffington Post, 30 January 2015)

Global Editors Network

The Global Editors Network (GEN) is a community committed to sustainable journalism and media innovation. GEN runs different programmes: Editors Lab, Startups for News, and the Data Journalism Awards.

Thanks to Nicolas Magand.

Emilie Kodjo

Written by

UN Communications consultant, Former Director of Communications and Public Affairs, The Global Editors Network

Global Editors Network

The Global Editors Network (GEN) is a community committed to sustainable journalism and media innovation. GEN runs different programmes: Editors Lab, Startups for News, and the Data Journalism Awards.