DOI: 10.60625/risj-0s9w-z770

Executive summary

Q5. To what extent, if at all, is your company worried about declines in referral traffic from big social media sites such as Facebook and X/Twitter? Base: 302.

The big social media reset

Underpinning the recent decline in ‘traditional’ social networks such as Facebook and X are two key shifts. First, the toxic nature of many conversations about news and politics have pushed many people towards private spaces such as messaging apps. This has led to even more unrepresentative and problematic content for the more open platforms to deal with.

Secondly, we have seen the strengthening of content-based networks such as YouTube and TikTok, where creators have access to increasingly powerful creation and distribution tools – and where the social aspects are important but often secondary. The Reuters Institute Digital News Report 2023 showed that these video-based networks are growing extremely fast across the world and are increasingly how younger audiences access their news.

This is challenging for many news publishers, partly because video is not their natural skillset, and partly because they offer few opportunities to link back to news websites where they can monetise content. Many traditional news organisations also struggle to get visibility, compared with young creators who are versed in the language and conventions of the platform.

Dylan Page, a young news creator from the UK, for example, has more followers and regular video views than the BBC or New York Times combined, even on important stories like Gaza (see below).

Publishers know that if they wish to engage next-generation audiences, they will need to up their game in short-form video and this is evident in our survey results with more effort being planned for both TikTok (+55 net score) and YouTube (+44).

But it won’t only be about consumption in video networks. Publishers also think there are opportunities to drive more referrals from messaging services like WhatsApp (+61), business networks like LinkedIn (+41), as well as relatively new sources of traffic such as Google Discover. With the launch of Threads and Bluesky, the job of audience engagement teams has become more demanding than ever.

What might happen this year?

Publishers focus more on direct traffic

In our survey the majority of publisher respondents (77%) say they will work harder on building direct links with consumers via websites, apps, newsletters, and podcasts – channels over which they have more control.

Around a fifth (22%) say they will be cutting costs. Worryingly, almost one in 15 say they have no strategy for this. ‘To be completely honest, we watch this helplessly and try to quickly pick up the last crumbs that fall off’, says one leading European news executive.

With less focus on social channels a key question will be how to attract audiences. Some publishers and broadcasters will look to spend more on marketing (17%); others will explore or put more effort into alternative platforms (20%).

‘[We] will prioritise our resources on our own channels and platforms where we can reach a young audience’, says Gard Steiro, Editor-in-Chief at Norwegian tabloid Verdens Gang (VG). ‘We are particularly focused on maintaining our strong position on Snapchat and building a loyal audience on TikTok. In addition, we believe that – for our part – there is untapped potential in optimising our content [better] for search engines.’

WhatsApp becomes a much bigger deal for news

Publishers say they will be putting more effort into WhatsApp this year following the launch of functionality that allows a range of personalities and brands to create broadcast channels. Users can follow or subscribe to a brand, react to posts using emojis, and forward them into private chats but there will be no open commenting.

Some early adopting publishers say that this new channel is already producing more referrals than Facebook and X. WhatsApp has the advantage of being widely used by mainstream audiences, according to the digital director of one leading French publisher. ‘There is greater trust in this network and the creation of communities is one of our key priorities.’ The New York Times has already amassed millions of followers and the Daily Mail has launched a number of channels including one that delivers daily news about the Kardashians. The publisher of one European digital start up believes these channels could, over time, have the ‘same power as push notifications and newsletters’. WhatsApp channels have been extended to 150 countries and similar functionality is also available in Instagram.

Threads could play a bigger role in news this year

Meta’s new short-form messaging service was billed as an alternative to Elon Musk’s X platform and quickly reached 100m active users. It has finally become available in the EU after complying with stricter rules. ^{Despite bosses saying that they don’t want the platform to be dominated by politics and hard news, much more of this is creeping onto the platform and, with elections on the way, we can expect more news brands and high-profile journalists to take the platform seriously this year.}

Social platforms lean into subscription

Big tech is not immune to the advertising downturn and has also been looking to pick up paid content revenue. Following Elon Musk’s decision to charge for blue ticks, other platforms have been offering extra privileges for paying members including, in some cases, preferential distribution. In a separate development Meta will be offering a paid version of both Facebook and Instagram in EU countries, costing an initial €9.99 a month, though this is principally a way of meeting privacy requirements set by European data laws, rather than a strategic shift away from a core advertising model. ^{Many people have complained bitterly about Facebook’s products tracking you across the internet. This year we’ll find out how much people value the ability to opt out.}

3. The future of search in the era of artificial intelligence

Looking to the year ahead, publishers are worried about a further substantial reduction in referral traffic as AI becomes integrated into search engines and other gateways. Google, Microsoft, and others have been exploring new ways of surfacing content known as Search Generative Experiences (SGE), that serve direct answers to queries rather than provide the traditional list of links to websites.

Microsoft Bing search was the among first to integrate real-time news, through its collaboration with OpenAI, the creator of ChatGPT. The underlying models were trained on content from all over the internet including news websites. Now these capabilities are being integrated throughout the company’s extensive product set, rebranded as Microsoft’s Copilot AI assistant.

Google, for its part, has upped its underlying capabilities with the release of Gemini, a set of multimodal large language models with similar capabilities to OpenAI. Gemini, also trained on content from the internet as well as other sources, can operate across text, images, audio, and video and will power Google’s Bard chatbot and all of its search generative experiences. Google has already rolled out an experimental version of AI-generated search in more than 100 countries (see example below) which can be triggered by text or voice queries. Google has been careful to include links to websites from which content has been drawn, but there are fewer of them and publishers worry that there will be even less reason to click on them.

AI disruption is going to increase business model problems. Search Generative Experience is going to replace some of the media outlets’ functions.

Chani Guyot, Publisher, RED/ACCIÓN, Argentina

But it is not just internet search. We’ll see a proliferation of conversational AI assistants this year built into computers, mobile phones, and even cars that will change the way we discover content of all types. Perplexity.ai is a website and browser extension that allows you to ask questions or summarise any news article. Pi.ai bills itself as a companion capable of coherently conversing on any topic including the latest news.

Public awareness of these tools is growing, according to a YouGov audience survey from the UK commissioned for this report. The poll shows that around two-thirds (65%) of the adult UK population have heard of one of five chatbots, with ChatGPT (59%) the most recognised example. Regular (weekly) usage of ChatGPT is still relatively low at just 9%, though this rises to 25% for the 18–24 age group. Top reasons to use chat bots include asking quick factual questions (37%), writing essays (29%), and researching a topic of interest (28%) – though many users are still just experimenting to see what it can and can’t do, just like publishers.

Finding the latest news (8%) is not currently a big draw but this could change as big companies add better real-time capabilities this year. Much will depend on how the search providers, particularly Google, which accounts for around 90% of the market, decides to proceed. It needs to balance its existing business model, which is built on links, with the need to remain competitive with other providers.

What might happen this year?

Copyright battles ahead

This year will see growing pressure from publishers to compensate for any loss of traffic and revenue resulting from AI powerhouses such as OpenAI and Google training their systems on historic data or delivering real-time news.

The New York Times’ lawsuit against OpenAI over alleged copyright infringement is likely to be the first of many this year. The lawsuit claims that ‘millions’ of Times articles were used without its permission to train the system, that ChatGPT sometimes generates ‘verbatim excerpts’ from these articles, and that the tool is now competing with the newspaper as a trustworthy information source.

Work in progress research by the Reuters Institute suggests that by the end of 2023 roughly half of top publishers had stopped allowing big AI platforms to access their content. This ranges from almost 80% of top publishers in the United States to less than 20% in Mexico. ^{By removing permission, publishers are in a better position to make deals with the platforms.}

A recent landmark agreement between Axel Springer and OpenAI, said to be worth tens of millions of euros a year for content drawn from Bild, Politico and Business Insider is seen by some as a possible model for the wider industry. It splits the agreement into a flat fee for historic data with an ongoing annual fee for new content, plus extra payments for content that is most frequently used by AI. ^{An earlier deal with the news agency AP also provides access to their historic archive and new stories in return for access to technology and expertise. Publishers hope that any arrangements will reward high-quality news and information from reliable sources.}

‘There is an opportunity for the industry to work with AI players to design a symbiotic ecosystem and that’s an opportunity we must not squander,’ says the COO of a leading UK news provider. ‘We know what happened the last time.’

If these voluntary arrangements fail, some news organisations will use their influence with politicians to push for an extension to arrangements already in place in a number of countries including Australia and Canada that require the biggest platforms to negotiate payments in return for surfacing news links.

Money for deals may not be equally shared

Despite some progress, most publishers in our survey are not optimistic that this new phase of negotiations will work out well for the news industry as a whole. More than a third (35%) of respondents felt that only a few big media companies would benefit while around half (48%) predicted that at the end of the day there would little money for any publisher.

4. The business of journalism faces further upheaval

Continuing volatility in the advertising market and further audience shifts towards digital will make this another tough year from a business perspective. In the United States, around 20,000 media jobs – not just news – were lost in 2023, about six times the number in the previous year. Between two and three local newspaper titles close each week in the United States, according to a recent report, ^{leaving huge swathes of the country as so-called ‘news deserts’. Other titles survive as ‘ghost newsrooms’ relying on shared or synthetic content with not a single permanent reporter on the payroll. TV news channels, which have significantly underinvested in digital, are also under pressure as audience attention switches to streamers. Meanwhile digital-born brands, such as Vox and Vice, that have built businesses via social platforms are retrenching.}

Amongst a wide range of publishers that have shifted towards paid models, including upmarket national titles and niche brands, there is a more positive outlook. Our survey shows that the publishers included in our survey reported continued subscription growth last year, despite the difficult economic situation. Even so, in many cases, increased revenue from digital has not been enough to make up for decline in print and advertising revenues. ‘Digital growth is stagnating, while print is plummeting’, is a typical story heard from national and regional publishers across Europe.

If they are to survive, media companies will need to wholeheartedly embrace digital, while continuing to squeeze revenue from traditional channels. In many parts of the world, however, print distribution networks are so weakened that we may be close to reaching a point where regular home delivery becomes unviable.

Looking at the most important future facing revenue sources across publishers, digital subscription and membership (80%) has extended its lead over display advertising (72%) as the most important revenue focus for the year ahead, with most publishers looking to combine three or four different income streams. Both events and e-commerce have increased in importance, but there is less hope than in previous years for extra income from big tech companies (20%). Some of this relates to Facebook winding down its payments to news publishers as current contracts end, while potential payments from AI platforms remain uncertain.

What might happen this year?

With subscription growth slowing, most heavy news users already signed up, and churn becoming an increasing burden, we can expect a slew of product and pricing innovations this year:

It’s all about the bundle

Over the last year, the New York Times has been moving as many of its customers as possible to the ‘all access’ bundle, which includes NYT Audio, The Athletic (Sport), Wirecutter (reviews), Cooking and Games. CFO Will Bardeen notes that ‘bundled subscribers retain and monetize better than news-only subscribers’ and the Times has an aim to get 50% of its 10m subscribers to take up the offer. ^{Games like Connections, used by around 10m users a week, are designed to build habit and affinity as well as drive referrals to news. Over time, all-access subscribers may find it harder to leave and the price can be increased.}

Northern European publishers such as Schibsted, Amedia, Bonnier, and DPG Media are also embracing the bundle, albeit in a slightly different way. Using their market strength, they are combining local and national publications with magazines and paid podcasts into a one-stop-shop offering.

So far most of these bundles have been intra-company collaborations, but in 2024 we can expect to see some publishers partnering with other providers to add extra value to their bundle.

Cheaper product extensions

We’re likely to see more upmarket news brands offering ‘lite’ versions this year in an attempt to woo audiences – often younger – that are interested in news but are unwilling or unable to pay the full price. This trend was started some years ago by the Economist’s Espresso, followed by the FT, which offers eight curated articles a day for £4.99 a month (FT Edit). Other lower-cost products in the pipeline include audio first news editions and combined newsletters and podcasts built around niches such as parenting and sport.

Differential pricing

Until recently, most customers paid the same ‘sticker’ price for their news, but as we discovered in our recent report on digital subscriptions, ^{that is no longer the case. Once the initial trial discount has run out, publishers are charging a renewal price based on usage (what they think the consumer might pay), or consumers themselves are negotiating a price that works for them. Many of these techniques have come from telecoms, but with better data available they will be increasingly applied in the news industry this year.}

Subscription latecomers join the party

All eyes will be on the new CNN CEO Mark Thompson as he attempts to breathe new life into a cable network which has been losing both audiences and revenue. As well as refocusing the core brand, Thompson is expected to launch a number of digital subscription products – just as he did at the New York Times – as well as revamping the CNN Max service, which sits inside Warner Bros’ Discovery streaming service.

Google’s imminent phase out of third party cookies is pushing more publishers to consider paid models. Even MailOnline, one of the largest English-language websites in the world, feels the need to supplement scale-based advertising with a limited (Freemium) subscription model. In early 2024 it will start charging for around 10–15 premium articles a day in the UK. Opinion-led broadcaster GB News has already begun charging for some online articles along with events that provide access to the channel’s presenters.

5. Changing formats of news: more audio and video on the way

For the last 20 years, the canonical article has been at the centre of how news information flows on the internet have worked. These text-based articles have been easily searchable and linkable, conversation has coalesced around them, and they have also driven the lion’s share of monetisation. Other formats such as audio and video have mostly taken a back seat.

But in recent years this has started to change, partly because of new devices, partly because platforms have emerged that specialise in audio and video creation, discovery, and distribution, and partly because many younger consumers show a preference for using these formats over text.

Some publishers talk of this format shift as a ‘second phase’ in the digital revolution because for many legacy newsrooms, it will require a significant cultural shift away from text towards multimedia production. Our survey this year clearly shows the direction of travel, with most news organisations planning to produce more video, more podcasts (and more newsletters) in 2024 and broadly the same number of text articles.

This chart is also closely linked to monetisation shifts. With less referral traffic coming from tech platforms, formats that drive engagement and loyalty (audio and newsletters) – and therefore support subscription models – have become more valuable. The exception is short-form video, where investment is more of a response to the need to attract younger audiences.

What might happen this year in audio and video?

Vertical video as default

We can expect to see attempts to engage young people with news on platforms where they are spending time, such as TikTok and YouTube Shorts. These initiatives will mainly be about trying to build relationships rather than making money, but in 2024 publishers will increasingly look to bring these storytelling techniques back into their own news websites and apps. From a talent perspective, short-form video storytellers will remain in great demand in 2024.

Podcasts push for visual distribution

In recent years political podcasts have proliferated, feeding the interests of the most interested users, and offering politicians an in-depth opportunity to set out their views in a less confrontational format. Many of these podcasts are now filmed, which allows them to reach even bigger audiences via platforms such as YouTube. And increasingly producers have been creating highlight clips, often distributed via TikTok, X, and other platforms.

Independent podcasters such as Alastair Campbell and Rory Stewart have started to record live shows on YouTube, while broadcasters such as the BBC are also turning podcasts into TV (BBC’s Newscast). All this will further blur the line between political podcasts and traditional television talk shows as we head into election season on both sides of the Atlantic. As part of this trend former Fox News host anchor Tucker Carlson is to create his own video network for political chat and plans to sell subscriptions directly to the American people from April.

Paid podcasts

While podcasts have traditionally been free, we can expect to see a further move to premium and mixed models in 2024. With most big advertisers focused on the top 100 podcasts, those with smaller reach, including many news ones, are finding it harder to pick up revenue. Partly as a response, Substack, which is best known for helping monetise newsletters, is supporting more podcasters into subscription, often in combination with a newsletter.

The Economist has gone one stage further and launched an audio subscription (£4.90 a month or £49 a year) that sees the majority of its shows sitting behind a paywall. Only its daily news podcast, The Intelligence, has remained free. The move has partly been made possible by a new technical solution from Apple which makes it a bit easier to listen to subscription podcasts on their platform while other providers such as Acast have also introduced improved paid options. These are likely to encourage others to follow suit this year. The Economist’s podcasts are made by a team of around 30 people, now about 10% of its total editorial headcount.

Specialist newsletters for a price

Some bigger publishers such as the Financial Times are experimenting with new ways of charging for popular newsletters without having to pay for a full price subscription. Unhedged, which covers markets and companies, and Inside Politics are on sale for $7.99 and $4.99 a month respectively. In early 2024, the FT will be experimenting with a ‘limited run’ paid newsletter series called Sort Your Financial Life Out, with Claer Barrett potentially opening up a revenue stream for more educational newsletter products.

More widely, subscription-based decision-maker briefings, also delivered mostly by newsletters, are gaining traction in Europe. Süddeutsche Zeitung recently launched a series of newsletter ‘dossiers’ that aim to unpack complex issues around politics and technology for specialist audiences. Table Media and Pioneer Media are other examples of thriving specialist newsletter (and podcasts) aimed at professionals.

6. News disconnection and selective news avoidance

It is clear that there will be more video, more audio, and more newsletters in 2024, but while it makes sense for a publisher to lean into these areas of growth, additional formats may not be what some audiences are looking for. Our own research suggests news consumers often feel overwhelmed by the amount of choice they already have. Platform and business incentives continue to push towards maximising attention but this is often not respectful of people’s time. Our survey reflects this dilemma as news organisations try to balance the needs of business and the interests of news consumers.

Advertisement

Issues of news overload and fatigue, highlighted in our Digital News Report for more than five years, ^{have become a major preoccupation for news executives. Research shows that despite – or perhaps because of – the political and economic upheavals of the last few years, many consumers are engaging with the news less often or not at all. ‘Information fatigue is eating away at our reader’s willingness to read and pay for our content,’ says the editor of one European membership organisation.}

Many public media organisations worry most about news avoidance amongst the young, but others see the trend cutting across age groups: ‘News avoidance is on the rise in Switzerland, which we need to counter with new journalistic approaches, topics and formats,’ says Marc Isler, Chief Revenue Officer, Tamedia.

It’s really depressing – I worry that journalism as a business has been too slow to reorient and rethink its role and purpose. People don’t miss journalism. But journalism miss people.

Lea Korsgaard, Editor-in-Chief, Zetland, Denmark

In our survey media executives outlined different ways in which they will be trying to address the problem in the coming year. These include better ways of explaining complex issues (67%), delivering news that doesn’t just point out problems but offers potential solutions (44%), and producing more inspiring human stories (43%). As in previous years, we see that newsroom leaders are less keen on the idea of more positive news (21%) or jazzing up the news with more entertaining or fun approaches (18%).

Advertisement

What might happen this year?

Letting go of doomsday narratives on climate

Climate change is a subject so profoundly unsettling many have learned to turn away from news coverage in what Wolfgang Blau, Co-Founder of the Oxford Climate Journalism Network (OCJN)^{, calls ‘the new climate denial’. Approaches to address this, by providing a sense of hope and personal agency, will gather pace in 2024. Some newsrooms have been upping their specialist coverage of green tech (Bloomberg Green) or creating eco-tips and inspiration for greener living. Earthtopia’s TikTok round of ‘Good News’ is consistently their most engaged-with content. Irish public broadcaster RTE runs a positive climate series called ‘Climate Heroes’, showing how individuals and businesses are making a difference.}

 

Advertisement

 

More diverse reporting perspectives

With traditional war reporters unable to get in to Gaza, many of the most compelling eyewitness stories have been told by local residents, including a new generation familiar with tools like Instagram and TikTok. Plestia Alaqad was one local resident whose documentation of the struggles of daily life brought a personal touch that has often been missing from mainstream media.

Young activists and journalists are also beginning to shape the news in other parts of the world. Ankur Paliwal is an independent journalist who has launched Queerbeat, a website that aims to transform the narrative about LGBTQIA+ communities in India. And the mainstream media are also looking to expand the range of voices that cover stories. The BBC’s TikTok coverage, for example, is now partly anchored by younger reporters.

Tools to allow audiences to reformat the news

AI-based tools that change the language of news to improve relevance and understanding for particular audiences will be an increasing feature of the news landscape this year. Artifact is a social newsreading app that can summarise news in different styles. In the example below, it can take the text of a Guardian article and rewrite it ‘like I’m five’, adapt it in the style of Gen Z, or as a series of emojis. Bots, apps, and browser extensions with similar capabilities will spread rapidly in 2024.

Advertisement

7. Generative AI and the impact on the newsroom

Since the public launch of ChatGPT just over a year ago, newsrooms have been coming to terms with the implications of advances in AI – and generative AI in particular. While there are deep concerns about trust and about the protection of intellectual property, publishers can also see advantages in making their businesses more efficient and more relevant for audiences. In our survey, news executives highlight back-end automation tasks (56%) such as transcription and copyediting as a top priority, followed by recommender systems (37%), the creation of content (28%) with human oversight, and commercial uses (27%). Other potentially important applications include coding (25%), where some publishers say they have seen very large productivity gains, and newsgathering (22%) where AI may be used to support investigations or in fact-checking and verification.

Back-end automation has become much more important over time. Just 29% said these applications of AI were very important when we last asked this survey question two years ago. This extra focus is partly because large language models (LLMs) that have emerged since then offer so many new opportunities to speed up and optimise routine tasks in the newsroom.

How the news media have been experimenting so far

Advertisement

• Summarisation: A number of publications including the Aftonbladet (Sweden), Verdens Gang (VG) (Norway), and Helsingin Sanomat (Finland) have started to add bullet points to the top of their articles. Aftonbladet have found that these summaries, known as ‘Snabbversions’ (or ‘quick versions’ in English), have increased overall engagement. Young readers in particular, they say, are more likely to click through for extra information.
• Headline testing: A number of publications have been experimenting with AI-driven headlines optimised for search, which are then checked by editors.
• Copyediting, notetaking, and transcription: AI tools to assist in these processes are becoming more widespread in newsrooms, with some job losses as a result. In June Axel Springer CEO Mathias Dopfner told staff that jobs that could be automated, such as proofreaders and some editors, will ‘no longer exist as they do today’. Transcription tools in small languages have also significantly improved in the last year. VG in Norway has recently rolled out its Jojo service while other publishers are using Good Tape, initially developed for internal use by the Danish digital-born brand Zetland.
• Translation: Le Monde uses AI to assist in translating articles, enabling around 30 stories a day to appear in its English edition – far more than would have been possible before. AI helps with the initial translation, with several human checks to follow. The software has been customised to recognise Le Monde’s style book and spellings.
• Image generation: A number of publications such as the Kölner Stadt-Anzeiger (Germany) and DennikN (Slovakia) are using tools like Midjourney to create graphic illustrations around subjects like technology and cooking. This is one of the most widely used forms of automated content creation.
• Article generation: The German tabloid Express.de has created a virtual journalist called Klara Indernach (KI), that now writes more than 5% of published stories on a wide range of subjects. Human editors still decide which stories get covered and review every piece of content, but the writing, structure, and tone are effectively outsourced to AI. Beyond authoring, the curation and selection of the headlines and links that consumers see on the website is now also automated, leading to a significant boost in click-though rates according to CEO Thomas Schultz-Homberg.
• AI presenters and newsreaders: Radio Expres in Slovakia has cloned the voice of a popular presenter (Bára Hacsi) whose AI replica (Hacsiko) now covers the nightshift, complete with commentary about the music and news stories taken from the company’s news website (Seznam Zprávy). Two radio stations in the west of England use a synthetic voice to turn text into an hourly radio bulletin. The company say they wouldn’t have been able to afford to employ a human journalist and that that the voice is so realistic few people have been able to tell the difference.
• TV channel generation: Going one stage further is NewsGPT, an experimental 24-hour television service where all the stories and all the presenters are generated by AI, without any human intervention. A small disclaimer says content may contain inaccuracies or unexpected outputs. NewsGPT, which can be watched via YouTube, bills itself as delivering news ‘without human biases’. The company did not respond to our requests for interview and it is not clear how it sources its news or whether it pays for it. Channel.1 AI, based out of Los Angeles, plans to launch a personalised news service in 2024 that learns what stories a viewer wants to see and promises to deliver them in any language. Journalists worry about the implications of this kind of automation for jobs, and for trust in the news media.

These developments have led to criticism that parts of the industry may be moving too fast, by exposing AI-generated content to users without clear rules on labelling or understanding the risks around the technology making mistakes (so-called hallucinations).

During 2023, automatically generated articles by CNET were found to be riddled with errors and poorly signposted. Sports Illustrated (SI) was also found to have included product reviews from a third party that were alleged to be at least partly written by AI, without proper disclosure.

I am concerned about quality of news as more and more companies are producing automated content. The overload of information will get worse and will make users even more confused.

Marek Kopeć, Product Director, Ringier Axel Springer Polska

The most compelling user case for AI in newsrooms is in the automation of routine tasks performed by editors, such as adding tags of SEO metadata. We do not believe that AI is a substitute for reporting stories, which will continue to be done by journalists.

Ed Roussel, Head of Digital at The Times and The Sunday Times

Advertisement

In our survey we find that respondents attach a very different level of risk to the various uses of AI. Content creation is seen as posing by far the biggest danger (56%), followed by newsgathering (28%). By contrast, back-end automation (11%), distribution, and coding are considered lower risk. This explains the higher level of focus right now on these back-end efficiencies and more caution around robo-journalism of different kinds.

What might happen this year?

If 2023 was a year for coming to terms with generative AI, this will be the year when newsrooms fully embrace the technologies and incorporate them into workflows.

AI tools everywhere

Advertisement

This year we can expect to see the emergence of sophisticated AI assistants looking over journalists’ shoulders and suggesting ways of making stories more interesting, more accurate, or more relevant. Helsingin Sanomat’s ‘Hennibot’, which is already built into its content management system (CMS), is a sign of things to come. It points journalists to language that can be improved or suggests story enhancements such as the most appropriate link.

At Aftonbladet, a Youth Assistant creates timelines, fact boxes, and Q&As written in the brand’s tone, but with a brief to also ensure that ‘an 18-year-old’ can easily understand the content. From the Philippines, Rappler recently won an AI Journalism challenge for a similar set of tools called TLDR (derived from the online expression ‘too long; didn’t read’) that converts the publication’s stories into summaries, graphics, and videos.

New AI roles and guidelines in the newsroom

Expect to see the appointment of more senior editorial figures to coordinate AI activities and strategy this year. In our Changing Newsrooms 2023 report ^{we found that 16% had already done so, with a further 24% actively planning to do so. The New York Times recently appointed Zach Seward, previously with Quartz, as its first editorial director of Artificial Intelligence Initiatives. The job description includes establishing how Times journalism can benefit from generative AI technologies, help journalists work faster, and identify guidelines and principles for how the technologies should and should not be applied.}

---

• 16% already have a designated AI leader in the newsroom (24% are working on it)
• 16% have detailed AI guidelines in place (35% are working on it)

Source: Changing Newsrooms 2023 report.
Q11. To your knowledge, when it comes to generative AI, to what extent is your organisation implementing the following practices? Base: 130.

Another related approach being pursued by publishers such as Schibsted is to build AI Labs, to lead experimentation and coordinate learning across publications and countries. Cross-company collaboration will also become more important in 2024. ‘We have engaged in a cross-industry process to craft a common transparency standard for AI-products in media,’ says Olle Zachrison, Head of AI & News Strategy at Swedish Radio (SR). The Reporters Without Borders (RSF) coordinated charter on AI and journalism is another important initiative bringing together 16 organisations and thinkers to define ethics and principles that can be applied around the world.

Advertisement

8. Artificial intelligence, elections, and the news

With important elections due in more than 40 countries in 2024, concerns have been rising about how these new technologies might be used by politicians and political activists – as well as by possible external ‘bad actors’ looking to influence the results.

In the recent Argentinian campaign, AI-generated imagery was used to catch voters’ attention and to denigrate both major candidates. ^{The eventual winner was depicted as a bloodthirsty ghoul, while a fake video showed the other main contender floundering in the trenches as part of the cast of the First World War movie 1917. In Slovakia, faked audio recordings of a candidate discussing how to rig the election were released just days before a tight election and proved hard for fact-checkers to definitively counter.}

So far, these interventions do not seem to have played a significant role in election outcomes, but in a recent poll, almost six in ten adults (58%) in the United States thought that AI tools would be likely to increase the spread of false and misleading information during the upcoming campaign.

These tools can generate synthetic images, audio, and video in seconds, as well as micro-target specific audience groups at scale, but electoral regulation has yet to catch up with this threat. Big tech companies such as Meta, Google, and TikTok are upping their defences, requiring users and political parties to declare any use of AI when uploading content, boosting election integrity teams, and working with fact-checkers to identify fake content quickly. In one early example, Republican presidential candidate Ron DeSantis’s campaign released an ad that mixed real images of opponent Donald Trump with fake ones, where he appears to be hugging former chief medical adviser Anthony Fauci. The ad did not include any disclaimer. The Republican National Committee (RNC) also released a political ad that used generative AI to visualise its view of what another four years under Biden might look like.

UK politics suffered one of its first ‘deep fake’ moments in October, when an audio clip of opposition leader Keir Starmer apparently swearing at staffers went viral, gaining millions of hits on X even after it had been exposed as a false.

Advertisement

The Israel/Gaza conflict has also thrown up countless examples of fake, doctored, or stylised images being used for propaganda purposes – from supporters of both sides. Recent versions of tools like Midjourney can conjure up hyper-realistic and detailed images of any situation that are hard to distinguish from real-life photographs, using simple prompts.

While the supply of synthetic content is certain to soar in 2024, the impact on consumers is harder to predict. Many of the so-called ‘deep fakes’ have been relatively easy to spot so far and only a subset of this content is designed to manipulate or confuse audiences.

Even so, the biggest danger may not be that people ‘believe the lies but rather that nobody believes anything any longer’, as political philosopher Hannah Arendt wrote in the context of an earlier period of upheaval. The vast majority of our publisher respondents are pretty downbeat about the implications for trust.

‘The explosion of crap content definitely has the potential to shake the trust in media,’ says Christoph Zimmer, Head of Product at Der Spiegel in Germany, but he also argues this could eventually allow some news media to ‘differentiate ourselves even more clearly as a quality medium, and thus even strengthen our position’.

What might happen this year?

Advertisement

Regulators get tough

In Europe, the Digital Services Act (DSA) has made the biggest social media and technology platforms legally accountable for content they publish, including various types of hate speech, and interference with elections. Fighting disinformation, including pro-Russian propaganda, will be a key focus for the EU with critical elections in a number of member states.

Regulators will want to show they are serious and have the power to fine platforms (up to 6% of global revenue) or even ban them from the EU market altogether. None of that can be ruled out with all eyes focused on X (Twitter), where human moderation has been scaled down substantially over the last year, even as other big platforms have been strengthening defences in order to comply with the new regulation. In December, the EU launched formal proceedings against X over suspected breach of its obligations to counter illegal content and disinformation.

Widespread labelling on AI content this year

The EU’s impending AI Act aims to allay the risks still further, but compliance is unlikely to be mandatory before 2026 so it is starting with a voluntary code of practice on AI-driven disinformation. A key plank is to get platforms to detect and then label any use of AI. Meta requires political advertisers to flag when they have used AI or digital manipulation in adverts on Facebook, Instagram, or Threads and Google has announced a similar approach. TikTok doesn’t take political advertising but is insisting on transparency from creators. These guidelines are, however, largely based on honesty and there are also concerns that consumers will ignore labels or that they might even generate a backfire effect.

Advertisement

Integrating AI into fact-checking workflows

A further defence could come with tools that use AI to identify fakes and false claims more quickly. The UK’s leading fact-checking organisation, Full Fact, has integrated a range of AI techniques to help human fact-checkers with identifying new claims, including from real-time broadcasts, using speech to text transcription, matching these to previous fact-checks, as well as helping to verify or debunk claims.

The Newsroom, an AI start up, has also been developing technology to cross-reference texts and political claims. It identifies different groups that support or refute a particular claim, as well as highlighting any contradictions with what a politician has said in the past. It also offers reference material to help a journalist dig deeper into a claim.

9. New devices and interfaces

For the last few decades computers or smartphones have been most people’s gateways to the internet, drawing us down rabbit holes and sucking up more and more of our time. But designers think that, in the era of AI, there could be better ways of accessing information and experiences. In 2024 we’ll see the launch of new devices that aim to break our reliance on the smartphone, using interaction patterns such as voice commands, eye movements, and hand gestures. Many futurologists think that we are moving towards a world of ‘ambient computing’, where wearable devices process what we hear and see in real time, supported by virtual assistants that can answer questions and serve up contextual information in any language.

• Hearables: There are already around 200m smart speakers in people’s homes around the world. Mostly sold by Google and Amazon these are about to get an AI reboot making them useful for more than just playing music. Headphones and earpods are already ubiquitous and more of them are becoming enabled with new voice and gesture prompts.
• Pins and pendants: Humane’s new artificial intelligence lapel pin ($699 + data subscription) has a mission to liberate the world from smartphone scrolling. ^{This small device magnetically attaches to your clothes and can be controlled by voice commands, by tapping a small pad, or by projecting a laser display onto the palm of your hand.}
• Smart glasses: Although Google Glass didn’t pan out, the technology is getting a reboot this year. Apple’s mixed reality headset, the Vision Pro, is due to launch in the spring or summer, initially in the United States, and claims to seamlessly layer digital content with your physical space. Though pre-release reviews have been positive, its high price point ($3,500) will not make it an immediate mass-market hit. The initial release is aimed at developers and Apple super-fans. More practical, lower price AR glasses are still a few years away.
• VR headsets: With lighter headsets boasting higher resolutions and broader field-of-views, the user experience of these devices has improved markedly over the last few years, yet immersive metaverse experiences have remained largely in the realm of science fiction or confined to gaming. But true believers like Meta’s CEO Mark Zuckerberg remain committed for the long term and the company will be revamping its Oculus headset range this year.

New devices shipping in 2024

In this year’s survey many of our respondents remain deeply sceptical about the idea of smartphone replacement. Four in ten (41%), however, do think that smart speakers and headphones will become a more important interface in the medium term, which partly explains the extra focus on audio we discussed earlier. Journalistic uses for both augmented and virtual reality are still hard to identify.

Advertisement

What might happen this year?

New smartphone killers in development

Though some are predicting that the first version of the Humane pin will flop, ^{similar devices could emerge this year. OpenAI boss Sam Altman and former Apple Chief Design Officer Jony Ive are reported to be collaborating on an AI-based device that is less reliant on screens, along with Softbank’s CEO Masayoshi Son. Ive has talked in the past about how product designers have a responsibility to try to mitigate what he sees as the addictive nature of technology.}

Chatbots with personality

If chatbots can seem a bit dry and boring, this year is likely to see an injection of a bit more sparkle. Meta recently announced a series of AI assistants based on US celebrities such as social media influencer Charli D’Amelio and American footballer Tom Brady. These bots, built with the full cooperation of the stars involved, are themed around passions such as sport, dancing, and gaming but can answer questions on a wide range of topics using a range of generative animations. Designed to make chatbots more fun, these AI celebrities will appear across multiple platforms including the metaverse. Publishers are already experimenting with news bots based on custom data, so adding the face and personality of a trusted journalist could be the next step.

Advertisement

Big year for spatial computing

The launch of the Apple Vision Pro, with a developer focus, will start to showcase the potential for these technologies, including some journalistic use cases. The New York Times is one publisher to have spent time creating 3D and immersive experiences in the past, including an exploration of the costumes of David Bowie and Nasa’s Insight mission to Mars. But up until now this has been laborious work.

New tools are now making the creation process quicker and new standards such as Web XR allow content to be published once but accessed across devices from a watch to a mobile phone to an Apple Vision Pro.

10. Conclusions

Technology continues to move at an extraordinary pace – well ahead of the capacity of most of us to absorb and make sense of. Publishers know that they need to embrace these next disruptions but the full implications are unlikely to be clear for some time.

Forward-thinking news organisations will be looking to build unique content and experiences that can’t be easily replicated by AI. These might include curating live news, deep analysis, human experiences that build connection, as well as longer audio and video formats that might be more defensible than text. But they’ll also be focused on using AI technologies to make their businesses more efficient in an increasingly difficult economic climate. At the same time, they will be working to package and distribute content in a way that makes journalism more relevant for different audiences, helping to address issues such as low engagement and selective news avoidance.

Advertisement

The impact of AI on the wider platform environment is harder to predict. Much will depend on emerging public attitudes to the technology but also on how responsibly the platforms themselves behave as the battle between accelerationists and ‘doomers’ plays out. Equally important to watch this year will be the outcome of legal cases around intellectual property which could open up – or severely restrict – the way news content can be used for training AI models without proper compensation.

Some news executives quoted in this report hold out the hope that an impending flood of unreliable synthetic content will strengthen journalism’s position and restore trust. Others worry that the public will lose trust in all information, further undermining democracies around the world. This year will give us more clues on which of these scenarios might play out as well as about governments’ appetite for regulating the power of the AI giants.

It’s just over a year since the public launch of ChatGPT and we’re still at the early stages of understanding what it means. That process continues but this will also be a year when we start to react and build a vision for journalism in the era of AI and how humans and machines can productively coexist.

Survey methodology

Advertisement

The 56 countries and territories represented in the survey included Australia, New Zealand, Taiwan, Hong Kong, Singapore, the Philippines, Thailand, Vietnam, Japan, Nigeria, South Africa, Kenya, Ecuador, Costa Rica, Nicaragua, Uruguay, Mexico, Brazil, Colombia, and Israel, but the majority came from the UK, US, or European countries such as Germany, Spain, France, Austria, Finland, Norway, Denmark, and the Netherlands, as well as Poland, Hungary, Slovakia, and Ukraine amongst others.

Participants filled out an online survey with specific questions around strategic and digital intent in 2024. Over 90% answered most questions although response rates vary. The majority contributed comments and ideas in open questions and some of these are quoted with permission in this document.

The author is grateful for the input from a number of publishers, academics, and industry experts when preparing this report via background conversations and emails. These included Rasmus Kleis Nielsen, Director of the Reuters Institute for the Study of Journalism; Alexandra Borchardt, Consultant and Senior Research Associate at the Reuters Institute; Manjiri Carey, BBC News Labs; David Caswell, AI expert at Structured Stories; Sarah Ebner, Financial Times; Pedro Henriques, The Newsroom AI; Troels Jørgensen, Digital Director at Politiken; Esa Mäkinen, Helsingin Sanomat; Craig McCosker, ABC, Australia; Matt Navarra, Social Media consultant; Grzegorz Piechota, INMA; Damian Radcliffe, University of Oregon; Martin Schori, Aftonbladet; David Tvrdon, writer and digital strategist, Slovakia.

Footnotes

 

About the author

Nic Newman is Senior Research Associate at the Reuters Institute for the Study of Journalism, where he has been lead author of the annual Digital News Report since 2012. He is also a consultant on digital media, working actively with news companies on product, audience, and business strategies for digital transition. He has produced a media and journalism predictions report for the last 14 years. This is the eighth to be published by the Reuters Institute.

Advertisement

Nic was a founding member of the BBC News Website, leading international coverage as World Editor (1997–2001). As Head of Product Development (2001–10) he led digital teams, developing websites, mobile, and interactive TV applications for all BBC Journalism sites.

 

Acknowledgements

The author would like to thank 314 news leaders from 56 countries and territories, who responded to a survey around the key challenges and opportunities in the year ahead.

Respondents included 76 editors-in-chief, 65 CEOs or managing directors, and 53 heads of digital or innovation and came from some of the world’s leading traditional media companies as well as digital-born organisations (see breakdown at the end of the report).

Survey input and answers helped guide some of the themes in this report and data have been used throughout. Some direct quotes do not carry names or organisations, at the request of those contributors.

Advertisement

The author is particularly grateful to Richard Fletcher and the research team at the Reuters Institute for their ideas and suggestions, and to a range of other experts and news executives who generously contributed their time in background interviews. Thanks also go to Alex Reid for input on the manuscript over the holiday season and keeping the publication on track.

As with many predictions reports there is a significant element of speculation, particularly around specifics and the report should be read bearing this in mind. Having said that, any mistakes – factual or otherwise – should be considered entirely the responsibility of the author.

Published by the Reuters Institute for the Study of Journalism with the support of the Google News Initiative (GNI).

This report can be reproduced under the Creative Commons licence CC BY. For more information please go to this link.