Media Relations: How to Work with AI-Driven Newsrooms

Media Relations: How to Work with AI-Driven Newsrooms
7. June 2018 Falk Rehkopf

Media Relations: How to Work with
AI-Driven Newsrooms

Media Relations: How to work with AI-driven newsrooms

For many, the word “newsroom” evokes an image of reporters sitting in an office and typing furiously away at their computers. But as newsrooms are embracing Artificial Intelligence (AI) and switching up their workflows, reporters suddenly aren’t the only ones filing and publishing stories – automated bots are doing the same as well.

In the past few years, newsrooms have made great strides in expanding their reporting capabilities, all through AI and other new technologies. Now that the landscape has changed, what does this mean for PR professionals and how does this affect the way they build relationships and pitch stories to reporters? And, are there any limitations of AI in the newsroom and how will its functions evolve in the next few years? In this article, we explore all that and more!

How newsrooms use AI

#1 – Suggesting content to readers

This year, Netflix moved on from using algorithms on designated batches of users to an image-based real-time algorithm to make personal recommendations on what movies and TV shows to watch. Amazon‘s algorithm recommends you items you might like, based on what the company calls ‘item-based collaborative filtering‘. Now, news portals and platforms are getting in on the fun as well.

By using AI technology, news portals can now offer recommendations for their readers and cater to them on a more personal level. In the UK, for example, The Times and Sunday Times are currently developing a recommendation service called James. This tool uses advanced algorithms to learn about individual preferences and personalises recommendations to readers in terms of format, time and frequency.

Miles ahead is China’s news app, Toutiao. When a reader logs in to the app, their news feed is updated based on their reading preferences, the average time they spend on an article and their location. Every day, Toutiao adds more than 200,000 articles from more than 4,000 news outlets serving more than 700 million registered users of which about 120 million log in every day, with an average engagement of 74 minutes spent on the app per day – that’s very impressive.

Toutiao - Nachrichten und Content-Plattform aus China

#2 – Monitoring the news

Monitoring the news is a highly mundane and tedious task – yet it’s essential for reporters who want to churn out relevant stories at breakneck speed and stay on top of the game. Thankfully, with the advent of AI technology, reporters can now outsource their monitoring activities to automated tools.

One prime example is Bloomberg, which has been using machine learning-enabled feeds to monitor breaking news for a few years now. Take Twitter, for instance. According to Bloomberg, perhaps 500,000 out of the 500 million tweets sent every day are relevant to its reporters. It would take countless man-hours to sift through every tweet and identify the important ones manually, so Bloomberg decided to enlist the help of technology.

Other AI tools that help companies monitor news include Ubermetrics’ own solution Delta. Another 94 tools, curated by the UK’s CIPR (Chartered Institute of Public Relations), can be found here.

#3 – Writing stories

If you told us that AI bots would be writing news stories 10 years ago, we probably would have not believed you. But hey, here we are!

There are tons of organisations who are now using bots to churn out news stories, including Associated Press (AP), which is using the Wordsmith platform to create publishable stories on corporate earnings. AP configured the tool to write in AP style and provided them with a story template to use and, in return, Wordsmith started producing publishable earnings reports.

Another example is Press Association (PA) in London, which is also using bots to churn out new stories. Here’s what’s impressive: unlike what AP is doing, the stories that PA’s bots are producing don’t fit into a certain category. According to editor-in-chief Peter Clifton, stories about smoking during pregnancy, recycling rates or cancelled operations – all written by bots – have been published online and in print (see example article to the left). Would you have noticed this was written by a bot?

Potential of AI in newsrooms

We’ve established the fact that AI tools are highly effective in applications such as recommending stories to users, monitoring news and even producing content, but beyond that, where does the potential of AI lie? And, given that AI is increasingly being seen as the next big thing, will it live up to all the hype?

For those who are wondering if AI can eventually replace journalists and reporters altogether, the answer is – no, we don’t see that happening, at least not in the near future. Among those who have spoken up to warn us against putting too much stock in AI is Sasha Koren, former editorial leader of the Guardian’s Mobile Innovation Lab.

According to Koren, the AI applications which we have in place these days just don’t match up to expectations. One particular use case she referred to was chatbots, which she found an “underwhelming experience”. While many bot developers promise that their products are able to “chat with you as if they are human,” Koren notes that all these bots do is query a database, which results in mechanical-sound answers, learned without any reflection.

Potential of AI outside newsrooms

Another group of stakeholders PR pros work closely with are bloggers – and these individuals can also benefit from using AI technology to streamline their workflows.

For bloggers, one of the most crucial steps of the process is content ideation – where you sit down and figure out the best topics to blog about. While bloggers had to spend hours on brainstorming in the past, the AI tools we now have make content ideation easier. Other than using AI tools to identify gaps in content and blog headlines which are likely to be well-received, bloggers can also use AI grammar-checking tools (such as Grammarly) to fine-tune their writing.

Other examples include Marketo‘s ContentAI or Uberflip which enable marketers and brand managers to show highly personalised content.

What if AI fails?

While we see more and more companies trying to sell tailored AI products to newsrooms, the fact remains that AI technology – applied in the context of the journalism industry – is still pretty new. Take into consideration the fact that developers of these AI products tend to be unfamiliar with the characteristics of the journalism industry and the standards that reporters have to adhere to, we foresee trouble brewing.

For print companies who are looking into investing into AI, it’s important to understand the full capabilities of these AI tools and ensure that the companies selling these tools live up to their claims. If the sales rep you’re speaking to assumes that your input data matches the data their developers use to train their algorithm and you don’t clarify this, you might end up with a product that doesn’t match your expectations. Because these tools are often proprietary to whatever company builds them, you probably won’t have access to the algorithm itself, which means that you might be stuck in a dead end.

As an extreme example, consider Norman, a psychopathic AI developed by MIT’s Media Lab and named after Alfred Hitchcock’s Norman Bates from his classic horror film Psycho. Norman was trained – being given data – by being shown horrific pictures of people dying to understand the impact on the AI. As a result, Norman was incapable of seeing anything but death and destruction even when looking at other pictures accentuating the risk potential, either on the data or algorithm side.

MIT Media Lab's 'Norman' - world's first psychopath AI

AI and ethics

AI technology comes with a distinct set of ethical challenges which affect everyone. A few years back, computer scientists from Carnegie Mellon University identified a bias within the Google ads system where women were six times less likely than men to be shown ads for highly paid jobs.

Embracing AI in newsrooms also opens up a can of worms when it comes to ethics. For example: what’s to stop AI bots from producing stories from data that it gathers in invasive ways? While reporters and journalists adhere to certain “codes” and use their own discretion to decide whether they should include certain (perhaps sensitive) nuggets of information in a news story, AI bots do not have the same capacity to do so. While there aren’t any workarounds to this problem as of now, AI developers and companies selling or using AI tools will have to assess these risks and come up with potential solutions in due time.

Interacting with AI-driven newsrooms as a PR pro

AI will impact PR workflows themselves but also the way PR pros interact with journalists and newsrooms. To get ready for a world where newsrooms increasingly work with AI, here are some great tips PR pros should follow:

#1 – Skill up for AI

A recent report by the CIPR predicts that 38% of PR skills will be either replaced or optimised by AI technology within the next five years. But, whatever PR task where AI can be of support, humans will still be needed for ‘intervention, editing, sensitivity, emotional intelligence, applying good judgment and ethics‘ the report concludes. First and foremost, PR pros need to embrace new technologies – such as AI, Blockchain and others – and understand how they work as well as the impact and opportunities these present for their industry.

#2 – Optimise content for AI algorithms

Today, optimising content for SEO purposes comes natural to most people working with content. A similar effort will be required to produce content optimised for AIs. With AI bots writing more and more news articles, PR pros will have to learn how to produce and provide content in a way that it can be picked up by bots.

#3 – Check your facts

Now that newsrooms are incorporating AI into their workflows, PR pros who deal with reporters and the media on a regular basis should ensure that all data and statistics included in press releases are up-to-date and accurate. With AI tools at their disposal, it’s easier than ever for newsrooms to verify the accuracy of these facts and figures and it’ll be embarrassing (to say the least!) to be caught disseminating outdated or inaccurate information.

#4 – Take media pitching to the next level

To ensure relevance, newsrooms working with AI technology will define and focus on specifically targeted audiences – something PR pros have always done. The toolset PR pros work with in terms of media relations should mirror the analytics tools editors and publishers use to effectively inform how to pitch to media companies.

#5 – Create PR content repositories

Newsrooms are relying on accessing databases where their AIs look for relevant content. The AI in turn will look for patterns within the data – i.e. stories – and will turn findings into articles. Setting up databases with e.g. branded content and linking these to AIs is a logical next step.

#6 – Be more competent than AI

Today, AIs handle routine tasks such as writing up financial reports but are not able to cover more complex topics. Journalists will therefore look to PR pros to deliver expertise, data and context. Therefore, PR pros should become true experts in specific fields and adopt the principles of life-long learning from a communications and topical perspective.

#7 – Bots will not replace relationships

Effective PR pros always understood the importance of relationships and this will not change. Irrespective of how AI will impact journalistic and PR practices, people will continue to make decisions on editorial content as well as the depth and direction of reporting. As Ivan Ristic of Diffusion put it, ‘A bot can’t lay claim to emotional intelligence, a cornerstone of all PR work. Teams employing AI handling external communication would be wise to have plans to manage reputation should anything go awry,” he wrote. “Humans build trust with humans—not bots.’

AI in newsrooms: future outlook

While it might seem as though we’ve seen plenty of advancements in AI in the context of newsrooms in the past few years, we’re far from being done. According to experts, AI will continue to advance and further refinements in the arena of natural language generation (NLG) will impact the newsroom profoundly and in more ways than one. As a PR professional, your best bet is to keep up with these advancements and adapt your methods of communicating with reporters and journalists accordingly. By doing so, you can find and tell better stories to make sure that you contribute impactful content and value to your audience.

Chief Marketing Officer