Can AI do journalism?

The past few decades have felt like a neverending bloodbath in the journalism industry.

News outlets big and small have been gutted by lost revenue. Every few months, an outlet announces another round of newsroom layoffs. Trust in news organizations is down, and the people in charge don’t seem to know how to fix it all.

Now, there’s a shiny new tool that promises to “do more with less” and maybe bring the industry back to its former glory.

If you’ve been reading the A.I. Agenda, you probably know where this is headed. AI has a lot of potential to make newsrooms faster, stronger and smarter. AI can research and crunch numbers faster than humans and ease the mind-numbing labor of transcribing public meetings.

There’s a real risk that AI-generated news will become a cheap, ubiquitous replacement for real journalism — but can journalists’ core responsibility to ask tough questions, pick up on nuance and push back against power be outsourced to a machine?

Since generative AI came on the scene several years ago, the news industry has seen its share of scandals:

That’s a lot to take in. So today we’re exploring AI from the lens of a newsroom — including our own.

But first, let’s start by looking at all the places AI is currently used in journalism. AI is already a tireless research assistant, crunching massive datasets to uncover trends, drafting initial versions of stories (because even journalists love a good rough draft), whipping up visuals and videos and translating content.

In fact, a Thomson Reuters Foundation survey revealed that over 81% of 200 journalists across 70 countries are already using AI tools somewhere in their editorial pipeline.

We’ve been thinking a lot about AI lately, in part because we use a lot of AI features within Skywolf, our legislation-tracking service. An AI model summarizes all the bills introduced at the Arizona Legislature, and a veteran human bill summarizer reviews and improves that summary before it’s published. Then she trains the model so it can do a better job next time.

So when we show you a Skywolf bill summary, that’s already based on an AI-generated, human-reviewed snippet of work.

But where are the redlines? Where’s the balance between utilizing AI for better, faster, more accurate journalism, and just getting lazy and handing our responsibilities over to the robots?

The journalism industry hasn’t quite answered those questions yet.

While major publications have taken steps to address the ethics of AI, many of their policies are surprisingly vague and often feel like placeholders, rather than concrete guidelines.

And many of those policies were written years ago, when AI tools like ChatGPT were first coming onto the scene.

The Washington Post, for example, touts transparency about its AI use, but stops short of outlining exactly how or when disclosure occurs. It’s AI policy says:

The Associated Press treads cautiously, acknowledging AI's potential, yet hedging with terms like "experiment with caution" and "benefit the mission." The AP explicitly discourages using tools like ChatGPT for publishing, but there's no clarity on oversight or enforcement.

The New York Times and others proudly emphasize human expertise, essentially declaring "machines can't replace judgment," which sounds comforting but sidesteps the real issues of when and how to use AI.

Tech publication Wired offers one of the more detailed policies, which explicitly prohibits AI-generated — and AI-edited — content.

This week, the Agenda set out to draft our own AI ethics policy. But we quickly found that, like many of the big national outlets, we’re still not sure where we land.

So we’re trying a different approach…

In the spirit of experimentation, we did something a little bold: We fed an AI model a full AI-generated transcript of one of our recent team discussions about AI and asked AI to write this story.

What you’re reading now – this section up until the Other News portion of the email – is the AI-generated product of that experiment, a kind of meta moment where AI helps us talk about AI.1

Our conversation covered a lot of ground, from the practical ways we use AI to the bigger ethical and philosophical questions about its role in journalism.

Some of us, like Adi, are fully on board with AI. Adi uses it for deep research, automating mundane tasks, and even talking through new concepts daily. For him, AI is already a trusted assistant, something he interacts with as naturally as a colleague. He shared how advanced research tools and the deep research mode of ChatGPT have made Googling feel outdated. Adi even sees a future where AI could handle personalized news delivery – no more scrolling through newsletters or articles, just a tailored digest delivered by a digital assistant on your morning walk.

Others in the conversation expressed a more cautious approach. Joe, for example, admitted to using AI only occasionally, primarily as a thesaurus or to rephrase a tricky sentence. He shared a story about trying to use AI for research on voter initiatives in Tucson, only to find that it “made up whole cloth that were lies.” He’s not quite ready to trust AI to write his stories, and he wonders what value he would bring if a digital clone could replicate his voice.

Nicole also highlighted a key limitation of AI: its lack of empathy and emotional intelligence. She described situations where being in the room, observing body language, and understanding the subtext of a conversation were crucial to getting the story right. “If an AI were to just hear that conversation, I think it would just spit out, like, they talked about these bills, and this is what they both had to say,” she noted. Nicole emphasized that while AI might be helpful for more mundane tasks, the uniquely human ability to connect emotionally with sources remains irreplaceable.

Hank's views on AI in the newsroom are a mix of caution, excitement and pragmatism. He acknowledges the potential for AI to make tedious tasks more efficient, but he's wary of how it might erode the authenticity of journalism. He emphasized the importance of maintaining a human voice, humor and perspective, arguing that readers come to the Arizona Agenda for its distinct style and insight – something AI can't replicate. While he sees AI as a useful tool, Hank is clear that it should remain just that – a tool – and not a replacement for the human elements that make their work valuable.

Curt, meanwhile, took a more analytical approach. He focused on the ethical questions surrounding AI in journalism, asking what the red lines should be and whether there are tasks AI should never handle. Curt acknowledged the potential of AI to assist with first drafts and outlines but stressed the need for human oversight. He emphasized that while AI can help streamline our work, it should not become a replacement for the core human elements of reporting – like empathy, conscience, and the ability to interpret nuance.2

Our discussion also touched on some broader questions: What happens to news consumption if AI can read and summarize everything for us? Will we get our news from AI-driven voices rather than reading articles or emails?

Adi even floated the idea of a world where journalism is more about feeding AI the right information than about writing for a human audience. That idea is both intriguing and a little unsettling.

There’s a lot more to explore here, and we’d love your thoughts.

How do you feel about AI in journalism?

And would you have known this story was written by AI before we told you?

Let us know – it will definitely be a human reading your responses. Probably.

Let AI swipe right for you: Got no game? Low on Rizz? Don’t worry. Tinder’s new AI will help you pick a match, message and even flirt with people for you, the Guardian reports. It’s all part of “effective coaching for struggling users” who lack social skills, the company says. Among the many concerns is whether other users will know when they’re talking to a bot or a real human, as well as the awkward first dates that are sure to follow an AI-bot introduction.

Hamburger University might need a new course: The fast food giant McDonald’s is planning to use AI at 43,000 locations to cut wait times, the New York Post reported. McDonald’s executives say installing Internet-connected kitchen equipment, AI-enabled drive-throughs and AI-powered tools will be just the beginning. One of the next steps will be to take the type of AI that is used in facial recognition and adapt it to make sure food orders are accurate.

Watching wildfires: Arizona Public Service, the largest utility in the state, is using AI cameras to detect wildfires from as far away as 25 miles, KJZZ’s Bridget Dowd reports. The cameras run 24 hours a day in northern Arizona and constantly look for smoke, with a fire mitigation team monitoring them.

Green nuclear: As we pointed out last week, utility companies in Arizona (including APS) are looking to nuclear power to satisfy the huge energy needs of data centers and AI. But they’re small potatoes compared to what Google and Microsoft are planning. They’re trying to reduce greenhouse gas emissions, while supplying huge amounts of energy to their AI projects, CBS News reports. Microsoft wants to reopen Three Mile Island, where the worst nuclear energy disaster in U.S. history unfolded in 1979. Google is looking at the small modular reactors that are also popular among Arizona utility executives.

Battlefield tools: The U.S. military is developing an AI tool to help deliver pain management drugs on the battlefield, Task and Purpose reports. The goal is to make it easy for soldiers who don’t have medical training to pinpoint the spot on the wounded soldier where the needle should go. That will help soldiers deal with pain and stay mobile on battlefields.

Time for a PhD union?: An AI program in Cambridge, Massachusetts has been reading medical journals for the past two years. Now, the people who own it say it’s ready to be a “scientific superintelligence,” the New York Times reported. The company, Lila Sciences, already made novel antibodies to fight disease and developed new materials to pull carbon from the atmosphere. Experts say it remains to be seen whether this new AI will be a powerful tool, or replace research scientists outright.

If you want John Wayne or Maya Angelou to read your company’s quarterly report to you, then check out this tool from Eleven Labs.

Their reader is an AI application that lets you load documents, or give it web links, and have the tool read it to you in various human voices. This is one of the cleanest text-to-speech products out there!