Automated journalism creeps into newsrooms leaning on AI

The use of AI in journalism is gaining steam and, at this early stage, newsrooms are still looking at how the tools can be used to help reporters tell deeper stories.

Is this article written by a bot? In a few years, it might be hard to say one way or another. Artificial intelligence is being applied to many different industries, and the areas of news and journalism are certainly no exception.

In fact, automated journalism is already helping create news articles and enhance storytelling. The Washington Post reported last year that its own AI bot, known as Heliograf, published 850 stories entirely autonomously, primarily reporting on sports and the outcomes of regional political races.

The company the Motley Fool projects that 20% of business content related to shareholder reports, legal documents, market reports, press releases, and even some articles and white papers will be generated by AI this year. The application of AI in the highly time-sensitive, competitive and tight-margin business of news and journalism is starting to speed up.

Natural language content generation gains traction

Much of the content generated by AI and machine learning relates to information and news that is of public interest, but doesn't require a large number of paid human staff to cover.

AI is being put to good use generating weather reports, financial industry summarizations, coverage of highly regional or local news and events, sporting event summaries, and other information that involves numerical information. AI systems can quantify that information and turn it into natural language text that's human readable.

AI systems are also being used to generate breaking news content to bridge the gap until human reporters are able to get to the scene. Reuters, for example, is using AI to scour twitter feeds to find breaking news before it becomes headlines. In this way, valuable information is transmitted as soon as it's available.

As mentioned above, the Washington Post has been using an AI approach that it calls robot journalism. Known as Heliograf, the company's bot first debuted at the 2016 Summer Olympics in Rio de Janeiro. Heliograf automatically put together stories by interpreting sports data and structuring it narratively based on patterns it learned from analyzing historical Washington Post articles. The use of the system allowed continuous reporting and accurate medal counts even for contests that were thinly covered by reporting staff.

Yahoo Sports has also notably adopted this approach to cover high school and college games.

Augmenting journalistic process and workflow with AI

Automated journalism is helping newsrooms by streamlining the media workflow and reducing the time, cost and complexity of getting news and content out in a timely manner.

The New York Times is using an Augmented intelligence approach leveraging an AI-based technology known as Editor that sits alongside journalists and identifies key phrases, headlines and text details. The system can provide on-the-spot research, content suggestions, links, fact-checking, and supporting quotes or facts to help improve the overall quality of the piece. This helps to significantly reduce the research workload of reporters and enables them to turn out better quality content faster.

How AI is tackling the problem of fake news

Similarly, BBC News has a product for their BBC News Labs division called Juicer that follows hundreds of news sources, aggregates information and extracts text that can be used by reporters writing about those topics. In effect, it's an AI-enhanced meta search engine that stores information about organizations, people, locations and other items of interest so it can later be quickly located and sourced. Because of this, relevant content can be identified quickly and current news stories of interest can be connected with other content generated earlier.

Similarly, the Associated Press (AP) uses an automated journalism tool they built themselves called NewsWhip, which observes social media postings on Twitter, Facebook, LinkedIn, Pinterest, Instagram, Snapchat and other sources. The system then synthesizes this information to keep tabs on how users are engaging with AP content, to share the content and to otherwise add value.

AP is also using that information to see how its content performs against competitors and what new and emerging keywords, topics, influencers and trends might be useful in future news coverage to garner the most attention on social media.

Improving content quality

AI is helping improve the quality of the content we read. The year 2016 saw the rise of fake news and articles that were written to mislead rather than inform. News sites and social media outlets alike are using AI-based systems to detect word patterns that may indicate a fake news story. These articles are then flagged for human inspection to see if the facts are real or not.

Major news outlets are increasingly using these AI-based systems to prevent hoaxes or fake stories from being propagated and spread through their legitimate news channels. They're also using these systems to expose those fake news sources and call out intentional fraudulent behavior.

Many online media organizations value contributions and comments from their readers. However, comments and forums can rapidly get out of hand with toxic submissions and inappropriate or harassing comments. The New York Times and other outlets are using AI systems to automatically moderate reader comments, keeping an eye on the tone of the language, the words used and other information that fits into undesirable patterns.

Google's Jigsaw group developed the Perspective API tool that categorizes comments into levels of toxicity. Automated or human moderators can then filter through these categories to remove comments that don't meet acceptable guidelines. The tools can also help users to control which comments they see.

Engaging with readers

Automated journalism systems are also helping content producers better engage with their readers. Machine learning systems are being used to learn individual reader preferences and recommend additional articles or even paid content that the readers might also enjoy and find valuable.

Media company Quartz received a grant from the Knight Foundation to produce BotStudio, which enables users to answer questions and engage with other interactive content as a way to identify relevant topics.

Other media companies are using AI-enabled recommendation engines to provide the same benefits. The Guardian recently introduced a chatbot that interacts with its readers, enabling greater engagement with news content. These chatbots can provide additional content or context that is not available in a story or provide deeper interaction with relevant topics.

Other media and content companies are using voice assistants such as Amazon Alexa to provide instant news briefings and enable more conversational forms of interaction with their readers.

While the concept of using bots and AI to generate news stories and change the way we interact with content might make those working in the fields of journalism and news uneasy, the reality is that media organizations aren't looking to remove the human element from reporting. Rather, AI is enabling journalists to do their jobs more effectively, helping publications to cover areas they previously lacked the manpower to handle and giving journalists the time to cover the stories that really matter.

Dig Deeper on Enterprise applications of AI

Business Analytics
Data Management