AI has become the main topic of conversation in recent months. People want to know how it will affect their everyday lives, as well as the impact it will have on the industry that they work in. Journalism has also been looking to answer these questions, and at the recent Newsrewired conference, they had a panel dedicated to it.  

Peter Houston, co-host of the Media Voices podcast, was joined by Tshepo Tshabalala, manager of JournalismAI, Eleanor Warnock, deputy editor at Sifted, and Jane Barrett, global editor for media news strategy at Reuters, to discuss the big issues about this relatively new technology. 

Want to check out the key points from the whole conference? Read our overview here. 

Impact on the news process 

Many would believe that AI has only been around for the last six months or so, after all the hype that surrounded ChatGPT. However, it’s been with us for a lot longer and is already being used in various newsrooms for different purposes. Tshepo told us that it currently impacts three key areas – news gathering, news production, and news distribution. 

In news gathering, machine learning platforms can use text analysis. This enables them to sift through large amounts of documents and gather information or topics that are relevant – this can also be done for images.  

Technology, such as Peach, can help journalists look through archives to assist with the news production phase. This gives an understanding of what has been written before, the basis for the news and what to write about now. Machine learning platforms can also help to identify quotes. 

Finally, we have the news distribution side of things. AI already assists with recommender systems, where readers are given a suggested article that should be of interest. It’s also part of comment moderation and can flag toxic or problematic comments (which a human then needs to decide is justified or not). 

Cutting down time  

One of the major benefits that AI can bring to the newsroom is cutting down the time it takes to do the more mundane side of the job. One of the generative AI tools that Eleanor uses has helped cut down the time it takes to edit audio for a podcast from two-to-three hours to just one. And Sifted are working on paid for tools around analysing start-ups. This is something Sifted have done manually in the past but generative AI can use this information, plus scrape text from the website, to put some analysis together.  

At Reuters, Jane spoke about how they are using PLX to pick out data points from a set of results that they want to learn more about. The team have worked on it extensively and now have the error rate down to one that is lower than human beings. The machine can do the time-consuming work of looking through the data, and then the journalist can find where the story is from that. 

Reskilling and risks 

As with any new technology, there are always risks to it and an adjustment period needed. Jane told us that generative AI is doubling every 59 days, so it’s hard to put any firm guidelines in place now as it’s changing so rapidly. Reuters has shared guidance on using AI with its workforce and has asked the newsroom to be open-minded, skeptical, and verify. 

Eleanor has seen the problems there can be with AI when using it to generate images. When looking for pictures of start-up founders, the data set used was mainly featured white middle-aged men. A human will always be needed to make sure there isn’t any bias and that there is diversity. There are also legal risks, copyright risks, IP risks, and more to be aware of and checks will need to be made so that everything is up to journalistic standards.  

Overall, Tshepo feels that a cultural change will take place in newsrooms. Education and training around AI is needed, and journalists and editors need to learn about the platforms and how to use them. Jane hopes that this will be a faster process than the digital transformation that took place around fifteen years ago, and that the industry learns from past mistakes we can move into this area quickly, but safely. 

For more on the impact AI is already having on the media industry, check out insight from panelists at the Voices by Vuelio event on the subject, featuring Press Gazette associate editor and New Statesman media correspondent William Turvill, The Telegraph’s Helena Pozniak, and freelance writer Amelia Tait.