Article published in ibm.com
Artificial Intelligence (AI) is playing an increasingly important role in the media industry, automating repetitive tasks and helping journalists focus on more in-depth reporting. Well-known organizations such as Bloomberg, Reuters, and AP have begun using AI to improve their newsgathering, production, and audience engagement processes.
Using AI in news gathering
Leading news organizations have implemented advanced technologies to aid in data research and analysis. Bloomberg has created a large language model (LLM) trained on financial documents, which improves sentiment analysis, entity recognition, and news classification for economic topics. Semafor, on the other hand, has collaborated with Microsoft and OpenAI to develop a tool called Signals, which helps journalists search for news sources in different languages.
At the Associated Press (AP), AI is being used to automate company earnings reports, processing information from press releases and stock analysis. Meanwhile, the Financial Times and the Wall Street Journal are working on AI models that predict upcoming controversial topics, helping reporters better cover trending issues.
How AI is helping in news production
Many newsrooms have begun experimenting with AI tools to improve their news-writing and editing processes. The New York Times doesn’t use AI to write stories, but it does use it to process data for investigative reporting, provide audio versions of news stories, and recommend articles to readers. It may also use AI to suggest headlines, summarize articles, and translate stories from English to Spanish.
The Washington Post has partnered with Eleven Labs to provide audio versions of its news stories, while the BBC has experimented with automated tools to create video and audio content. Additionally, AI tools like Newtral and FactStream help with fact-checking by identifying statements that need to be checked.
AI and audience engagement
In addition to producing news, AI is also influencing how readers consume it. Early testing shows that AI-generated summaries can increase readership. For example, a public broadcaster in Norway and a newspaper in South Africa saw increases in readership thanks to the use of AI-generated summaries.
The Washington Post has developed an AI-powered chatbot that helps readers better understand climate science by providing summaries of information from its archives. The model also warns if there is not enough information for a given question and refers readers to the original articles for verification. The Financial Times has built a similar tool that answers subscribers’ questions on current topics, although some testing has found that the answers can contain inaccuracies.
AI is also used to optimize subscriptions, tailoring content to readers’ behaviors and increasing their engagement with newspapers. The right use of AI for content curation can help readers be exposed not only to the news they care about, but also to important issues they may not have previously considered.