AI is endangering journalism–yet another industry under threat because of its quick, cheap, human labor-free ability to generate content. In 2024, outlets like The Messenger and Pitchfork collapsed, while bigger household names like Business Insider, Time Magazine, and the Washington Post were just a few of the many laying off journalists en masse (more information here). The decline of physical newspaper consumption thanks to the rise of TV and news apps, alongside most advertising transitioning from newsprint to digital, have done huge damage to many news outlets’ revenue streams. The digital age is forcing print media to choose: be eradicated or redefine yourself to survive. Analog in all its forms seems to be on the way out, practically speaking. Here’s a number to quantify: An estimated 2.5 newspapers shut down for good each week in the United States. They’re disappearing fast, and this Northwestern study only predicts an increase–not that there were many newspapers to begin with. Half of all counties also only have one news outlet–there were never many to begin with. In essence, the older print models that newspapers are used to will not survive in the contemporary digital age. How will the news industry adjust in the face of its decline?
Only a few large news outlets have successfully digitized their subscriptions–NYT is one of them, having also branched out into games traditionally found in newspapers like crosswords and sudoku into mobile games that refresh daily. Different types of journalism have also evolved differently–investigative journalism has hugely benefited from the availability of big data; cable journalism has proliferated under streaming services and partially shifted towards comedy or inciting strong emotional reactions.
But NYT is one of the few news organizations to digitize so successfully. Most smaller, local newspapers have been forced to close their doors. Larger names like the Huffington Post have seen a dramatic decrease in profit. With digital journalism replacing its traditional predecessor, articles and news are not only being consumed through screens, they’re being written and distributed through the digital world, too. The shift into the digital sphere has been paired with an influx in the amount of information available–more for digitized news outlets to compete with in a limited amount of space.
This surplus has led to many social media users to just avoid non short-form content altogether for the sake of preserving cognitive load. Viewers also avoid objective reporting and seek out information from articles that support their existing biases, further centering “clickbait-y” content and articles. These, as the name suggests, get more clicks than their non-clickbait alternatives.
Digital journalism also offers a larger array of resources to support the written story
–videos, audio, and direct links to other news pages simply leave print photography in the dust. And given that the smartphone is so deeply embedded in our daily routines, it’s simply easier to read the day’s news through a touchscreen. Better yet, have a content creator summarize it for you through your earbuds as you go about your day. Let’s face it–there’s no turning back to the old ways anytime soon.
The New York Times claims that AI “is always used with human guidance and review”, according to its principles for the use of A.I.. Its applications range from sifting through the internet’s massive surplus of data for investigative reporting, to drafting summaries and headlines, or recommending articles. AI has stealthily crept into journalism– though it’s currently not used to write actual articles, but for more innocuous things like translating or generating audio for articles in the name of accessibility. The NYT also claims to not edit or alter its image media with AI in any way. It places heavy importance on ethics and emphasizing human review prior to releasing the final product. However, the NYT is under harsher scrutiny from the general public, since it’s such a large and well-known name. Thus, it’s only fair for one to assume that smaller publications might have fewer qualms about AI ethics, not to mention the NYT’s comparably liberal audience likely fuels its stance towards AI use in journalism.
Quality checks are not always completed: Politico has faced controversy a few times for typos made by AI summarization models, imposing AI use abruptly without informing the editorial team prior, and taking up digital real estate that could otherwise be filled by human-driven content. With Politico, we see the flip side of AI in journalism–-intransparency about its application and making the false assumption that it'll give perfect responses each time. For example, an instance where AI misspelled Kamala Harris’ mother’s name was suddenly taken down without explanation. Another time, Politico covered the 2024 vice presidential debate between JD Vance and Tim Waltz, directly transcribing Vance’s comments about “illegal immigrants” despite this voiding Politico’s editorial guidelines. Politico is no small name either-one can only hope that these kinds of errors will be penalized rather than normalized in the future.
It’s vital to keep in mind that not all media outlets will have transparent AI policies like the NYT. And just because one iterates that they value ethics now doesn’t mean that all outlets do–or even that one that used to will continue to do so in the future. AI in its current form remains relatively easy to identify; there are clear signs if you know what to look for. The clash between AI and journalism (in addition to other creative outlets like visual arts) will only deepen and grow more complex as AI models and their machine learning capabilities advance and the “AI slop” stylistic identifiers disappear. It will only grow more difficult to successfully and accurately gauge what was made by man, and what was made by machine. There seems to be a vacuum in the AI regulation sphere, for both how and if it can be used as well as how far we should take this golden goose technology.
Questions about the tradeoff for AI’s efficiency–“just because we can, should we?” Yes, having an AI model spawn your articles into being is more cost-efficient and faster, but the cost comes at the rising unemployment among journalists and the loss of that unreplicable, intangible, human quality to man-made writing. Also something to keep in mind as technology envelops more and more aspects of our daily lives. Stay wary, stay vigilant while reading–it could very well be a Trojan horse.