This article was originally published on Northwestern University’s Medill Local News Initiative website and is republished here with permission. The text was adapted from the introduction of Medill’s report on the impact of AI on local news models. Download a full PDF version of the report here.
Journalism has experienced its share of revolutions, from Johannes Gutenberg’s invention of the movable-type printing press in 15th-century Germany to the high-speed presses of the 19th century to the disruptions of radio, network television and cable television as primary sources for live, breaking news. Yet even as competition soared and the number of daily newspapers declined, local news thrived as an industry. In a 1990 Washington Journalism Review article, former Chicago Tribune editor James D. Squires called newspapers “the most profitable legal business in America.”
Then came the internet.
The World Wide Web opened to the public in 1991, with documents and other materials becoming available via web pages and searchable via web browsers. Newspapers including The New York Times, The Washington Post and Chicago Tribune launched their websites in 1996 and would come to wrestle with whether and how to charge for online content.
The internet made news organizations instantly accessible to audiences in their communities and beyond while offering journalists an array of new tools to do their jobs more effectively. It also helped decimate the local news business model.
As more and more consumers opted to get their news for free online, print circulation and advertising revenues plummeted, and paywalls, digital advertising and online subscriptions didn’t come close to compensating. News organizations found themselves reliant upon social media platforms they didn’t own and couldn’t control, with revenues they couldn’t capture. The 2023 State of Local News Report — released in November by the Medill School of Journalism, Media, Integrated Marketing Communications — concludes that by the end of 2024, the U.S. will have lost a third of its newspapers and almost two-thirds of its newspaper journalists since 2005.
The collapse of the mainstream news media’s financial model has affected more than just struggling journalists and those who consume their work. It may also represent a threat to democracy, creating vast news deserts and the opportunity for ill-intentioned players to fill the void with misinformation and disinformation. Given these high stakes, philanthropists, business professionals, politicians and others are dedicating time, energy and hundreds of millions of dollars to help sustain local news.
Amid this bleak landscape, it’s understandable that some in the industry view the next technology-driven revolution with trepidation. Generative artificial intelligence has had an explosive impact on journalism and the broader culture since Open AI publicly launched ChatGPT on November 30, 2022.
ChatGPT, which stands for Chat Generative Pre-Trained Transformer, is a large language model-based chatbot that can create content based on prompts fed into it. With more than 100 million users signing onto ChatGPT within its first two months on the market, the global financial firm UBS deemed it to be the fastest-growing consumer software application in history.
This project — based on interviews with journalists, academics, philanthropists and technology experts from around the globe — seeks to identify the potential perils and benefits of generative AI on local news business models and journalism itself. They include …
AI in Journalism: Benefits
- Improve news outlets’ responsiveness with customers and consumers
- Help news organizations operate more efficiently
- Allow for greater personalization of news and how it’s delivered
- Inspire news organizations to rethink how best to serve audiences
- Enable more effective targeting of marketing and advertising to consumers
- Create new tools to improve storytelling and to monetize content
- Free up journalists for more original enterprise reporting
AI in Journalism: Perils
- Disrupt an already-ailing business model
- Exacerbate the loss of web traffic as search engines train on news content and answer search queries directly
- Introduce errors into stories that damage credibility
- Lead to the replacement of journalists with machine-generated content
- The lack of guidelines around its use could prompt ethical problems
- Limited resources could hamper small news outlets from capitalizing on its use
- Accelerate the spread of misinformation and disinformation
Many people who practice or care about journalism fear that generative AI, with its ability to create content with little human involvement, could be the final nail in the local news coffin. Given how some chain owners have prioritized cost-cutting and profit-making over sustained journalistic quality, what is to stop them from replacing more reporters and editors with robots? Can news consumers be relied upon to discern between human-reported journalism and machine-generated content — and does it matter?
There are further dangers. AI, which has shown a propensity for mistakes early on, could be prone to spreading misinformation and disinformation, either by accident or design. And if Google can respond to reader queries by producing answers generated by AI and sourced without attribution from online news reports — instead of by offering links to the articles themselves — digital traffic to news sites may suffer a monumental blow.
In short, is generative AI an existential threat to journalism and local news models?
That question is a huge dark cloud looming over this game-changing technology. But there’s another way to look at it.
Generative AI is only the latest in a long line of technological advancements that, when used correctly, should make work more efficient and easier. After all, computer layout programs eliminated the need for T-squares, glue and a paste-up department to get newsprint on the page. Cellphones and email boosted communication among editors, reporters and readers.
Modems allowed writers to file stories remotely, and then high-speed cellular made it possible for pieces to be sent from anywhere. Search engines empowered journalists to conduct research via button clicks instead of courthouse and library visits and exhaustive phone calls. Spell-check programs help writers and editors catch typos and misspellings, and style-checking tools flag possible punctuation errors. Those latter two are forms of AI.
Few would argue that using spell-check represents a threat to journalism.
So, AI represents opportunity as well as peril. How can this tool help news organizations do better? What can robots do so humans can be freed up to improve their work?
Can AI help news outlets rethink what they should be doing? Can AI help local news organizations solve the business problems that have been vexing human professionals for so long? Can the industry win battles with Big Tech over the use of its content to train its own models and to respond to user queries, leading to a potentially catastrophic decline in web traffic referrals? Is it possible for the journalism world not only to avoid the mistakes made with the advent of the internet but to create a new, better reality that leads to a more informed populace, a healthier democracy, and a robust, sustainable local news profession?
“When the internet happened, we basically were in fear and denial,” said Tom Rosenstiel, Eleanor Merrill Scholar on the Future of Journalism, Professor of Practice, Philip Merrill College of Journalism, University of Maryland.
“Now we have another chance because this is as big as the internet.”
