InPublishing: Closing the AI gap: how small news publishers can benefit

Investment in Artificial Intelligence (AI) by private companies doubled during 2021: did yours? According to the 2022 AI Index Report of the Stanford Institute for Human-Centred Artificial Intelligence that measures global trends across sectors such as retail, health and security, AI has passed the novelty stage and is now integral for how businesses are planning for growth and development post-pandemic and beyond.

Jack Clark of the AI Index said, “2021 was the year that AI went from an emerging technology to a mature technology – we’re no longer dealing with a speculative part of scientific research, but instead something that has real-world impact, both positive and negative.”

According to an Oxford Reuters Journalism Institute report, AI is now recognised by industry executives as vital to media production and the business model. Of the companies they surveyed around the world, 85% said that AI will be important for content recommendation, 81% said it would drive newsroom automation and 69% said AI was critical on the business side to help attract and retain customers.

But for many newsrooms, AI is still an idea, not a reality. Small publishers especially are struggling to explore, let alone implement AI tools and systems. But help is at hand. In this article, I want to explain how local, specialist or ‘small’ media organisations can avoid missing out on the potential benefits of automation, machine learning and other ‘AI’ technologies.

Limited resources

When it comes to innovation and enterprise, small businesses can be nimble, novel and creative. But they are likely to have fewer resources such as time, money or expertise to devote to technological R&D. According to our research, that is very much the case in news media. The kind of technologies such as machine learning, natural language processing and automation that come under the umbrella term of ‘AI’ are seen as complex to understand and difficult to deploy. And to a degree, that can be true. Adopting AI does require knowledge, strategic planning and effort.

That has led to an inequality problem for publishers. Overall, the industry is not big enough to be able to do deep research to compete with the big technology companies like Google or Microsoft and research labs at universities like Stanford or Cambridge. So a lot of the technology is being developed for other industries. That means the news media often have to work harder to design and adapt AI for their own use cases.

Within the news industry itself, there is a gap between big companies such as the Wall Street Journal, Washington Post, BBC or Axel Springer who do have R&D departments and the rest of the industry where product development and technological development is much more ad hoc, if it happens at all.

So how can we bridge that gap and, first of all, is it even worth trying? For the last three years, my JournalismAI project based at the LSE has been working with a range of news organisations around the world. We did research, created online training programmes, and a case study resource hub. We have also run a series of collaborative projects where tech and editorial people from different newsrooms worked on finding AI solutions for challenges such as automation or personalisation. They found that when done responsibly and with planning, AI could make their journalism more efficient and effective. But one of the key lessons of this process was that smaller newsrooms often found it harder to implement AI solutions.

Missing out on opportunities

A recent research report by the Associated Press in the US showed that local news organisations are missing out on the benefits of AI. They did an in-depth study of 25 local newsrooms to judge their ‘AI readiness’. Some were using AI tools, but overall, US local news organisations were too busy trying to keep afloat to spend time on innovation. As one US local news executive told the researchers, “My biggest fear is: Can we move fast enough to keep up with new technologies? I know things are moving very quickly in AI, and I ask myself organizationally, are we going to be nimble enough to be able to take advantage of opportunities?”

The AP report did identify a whole series of possible uses for small publishers across all stages of the news process. The local news people told AP that they could use AI, for example, to do interview transcription as part of newsgathering. Or the automation of High School sports results. In content creation, AI could be useful in automatically generating material for social media. In news distribution, AI could create personalised newsletters or A/B test headlines. And on the business side, AI automation could help them with audience analytics, or automated advertising design. There are already tools and software systems that have been developed to address these issues. Clearly, when they took time to think it through, the news executives identified benefits and were keen to try. The problem, it seems, is finding the time for someone to get up to speed on AI so they can start the process that will lead to implementation.

AP is now working on a programme to help those US local newsrooms to catch up. But that leaves a lot of small news organisations around the world who are also struggling to get on board the AI train. One response by the LSE JournalismAI project was to create an ‘AI Academy for Small Newsrooms’ which we piloted last year. Twenty people from small newsrooms from Sweden to South Africa took an online course with two weekly sessions over two months where they were coached by AI experts on topics such as defining AI, practical use cases, skills and management, and risks. They all had to create their own ‘playbook for AI adoption’. The feedback was hugely positive so we’ll be running more academies this year.

The lessons we’ve learnt from this whole process is that small news organisations can benefit from AI technologies if they think it through. As one person who took part in our academy told us, “AI implementation in the newsroom is not all about the technology, there are other important aspects to factor in like the resources: time, money, people.”

Here were the key steps that we think small newsrooms should take when they consider using AI:

Make sure that you have someone responsible but make sure they have the attention of everyone in the organisation
Find out about the basic AI terms so you can avoid the BS and the hype
Think about the problems you would like to solve
Remember: AI might not be the right solution for your needs
Research what tools and systems are already out there
Talk to other organisations about what worked for them
Start small and be prepared to work differently – consider collaboration with other news organisations, universities or start-ups
Don’t expect easy wins or immediate big benefits

At JournalismAI, we have prepared an online ‘JournalismAI Starter Pack’ which is designed for anyone from any size of publisher to help them through the first steps and to plan a strategy. It has all the resources to get you on your way.

We don’t think that AI technologies can solve all of journalism’s problems. And they bring a few new problems in their wake such as algorithmic bias and dependency on tech providers. AI might not be relevant to your organisation. But the technology is developing rapidly. New products and systems for news work are being launched every week. If you don’t use it, a competitor might.

At least, take a few hours to check out the potential. As a senior executive at a major international news organisation told me recently, “The main challenge is whether the news media can take advantage of AI to benefit from the efficiency, insight and products that it can provide. If ‘we’ don’t then someone else will. There is a huge opportunity cost to not acting now.”

Don’t use size as an excuse. Yes, the big news organisations seem to be getting bigger. But around the world, there has also been an explosion of micro and medium-sized media that can service local or niche audiences and provide diversity and variety in publishing. Check out (https://bit.ly/3PldjZO) how the tiny team at Peru’s Ojo Público used AI to power their anti-corruption investigations. AI is already proving that it can turbo-charge human journalism and small news organisations do not need to miss out.