Why it’s about time we talk seriously about AI and journalism

“This is a complex technology and it is going to take years to discover its impact”

Last February Polis, the international journalism think-tank at the London School of Economics and Political Science, began a research project in collaboration with the Google News Initiative that aimed at investigating the intersection of journalism and artificial intelligence.

After seven months of work and collaboration with 71 news organisations currently using this technology, Journalism AI is about to publish their findings in a report that will be launched at Hacks/Hackers London on November 18.

Ahead of the event, we talk with the founding director of Polis, Charlie Beckett, to understand what AI really means for journalism.

Why this report? Why is it necessary to speak about AI now?

This report comes at a critical time for the news industry but also at a vital time for these AI technologies. News organisations around the world of all types are starting to use AI in significant way – from investigations, to personalisation, to the way news is consumed itself. Our report shows that there are some imaginative and efficient uses already in place, but there are also a host of challenges such as training and algorithmic bias. Journalism is under pressure as a business but it is also struggling to prove its value in a world of misinformation and political conflict. AI won’t solve all those problems, but unless we understand it better, we are going to miss a chance to improve journalism’s ability to thrive in a data-driven world.

How would you summarise the state of AI in journalism?

The state of AI in newsrooms ranges widely from early adopters to some that are still avoiding the technology. Most newsrooms are using AI in piecemeal ways and are only just starting to see how they can create a strategy to maximise its effectiveness across all aspects of the news process from newsgathering, to content production, to distribution. I think the best comparison is with the early stages of social media about ten years ago, when some news organisations were just experimenting with Facebook or Twitter, while others understood that this was eventually going to change their whole business and editorial systems.

AI could widen the gap between large and small media companies. How can we prevent it?

One of the biggest issues around AI for journalism is that the news industry is much smaller with fewer resources for research and development than other industries such as retail, security or health. And within the industry itself, bigger newsrooms have more human and financial resources to adopt this technology in a serious and systematic way. So the danger is that smaller news organisations might get left behind. However, there are also signs that local and specialist news organisations can use AI to get a comparative advantage if they focus on their needs and goals. It also looks like there are going to be opportunities for nimble, high-quality intermediary organisations who can create the niche tools or services that newsrooms are going to need as AI spreads through journalism production. There is going to be a lot more space for start-ups and collaboration.

Among the biggest challenges to adopting AI cited by your respondents, is that there is cultural resistance (24%) including the fear of losing jobs. Is this a real threat or a lack of knowledge of what AI is?

The cultural issues around AI adoption that our survey discovered are familiar to anyone who has been through previous technological changes in journalism. And it is right that we don’t abandon good journalism values in the race to use the bright, shiny new tech. But the people we spoke to said we need a massive investment in training, skills and education. Not just for specialist IT staff but for everyone in the newsroom. AI works best when it augments existing processes and that means that everyone is going to have to understand AI to some degree, from the top managers right through to marketing or editorial people.

How AI can help in the fight against misinformation?

AI is not going to kill off journalism, but neither is it going to save it by itself. However, it can play a part in helping journalism prove its value to the public at a time when misinformation and polarisation are making it tougher than ever for the citizen to find the facts and debate that can help them live their lives. The fact is that we are going to all be living in a world where huge flows of data will be driven by algorithms, often shaped by technology companies and other organisations. AI can help journalism to fight ‘fake news’ directly, for example through automated authentication systems. More generally, it can help show the public which information to trust and to create better ways to connect good content to the people who need it. Our report showed many innovative ways in which this is starting to happen. Above all, it involves making ethical and well-informed decisions about how AI will play a role in journalism. This is a complex technology and it is going to take years to discover its impact. But it is clear to me from our research that over the next few years, this is going to be a significant factor in how the news industry evolves. The struggle to get this right starts now.

Want to join Hacks/Hackers London? Check our website or subscribe to our newsletter for future events and info about our guest speakers. You can also watch the talks from previous meetups on our YouTube channel and follow us on Twitter and LinkedIn.

If you want to know more about Journalism AI, or if you want to make sure to receive the report next week, do not hesitate to get in touch with Mattia Peretti at M.Peretti@lse.ac.uk