HE COMMENTARY: Spare the Lecture — Newsrooms Have Used Writing Technology for Decades

AI Slop or real news? What about propaganda, sucking up to power, and press release printers calling themselves "news outlets?" Gag me!

Artificial intelligence isn't journalism, but neither was a typewriter. What matters is who does the reporting, checks the facts and takes responsibility for what gets published, and that there is newsworthy content. HE seeks to break the news. And this report by another "rag" was like listening to Chief Justice Rush comment on civil discourse, or civility, or Mitch Daniels discuss integrity or good government.

INDIANAPOLIS — Indiana Capital Chronicle recently examined how Indianapolis newsrooms are using artificial intelligence, joining a debate that sometimes makes it sound as though asking ChatGPT to help research a story is one keystroke away from turning the newsroom over to HAL 9000.  For the less learned HAL 9000 is the fictional artificial-intelligence computer in Stanley Kubrick's 1968 film 2001: A Space Odyssey, based on work by Arthur C. Clarke.

See their article which left HE out for some reason as we are the fastest growing media company in Indiana, since our Editor-in-Chief was released from federal prison, where he never belonged. https://indianacapitalchronicle.com/2026/05/20/how-indy-newsrooms-are-using-ai/

There are legitimate reasons for caution. AI can misunderstand information, invent facts and occasionally deliver a completely wrong answer with the confidence of a television pundit on election night. Nobody should publish something merely because a computer said it was true.

But irresponsible journalism wasn't invented by artificial intelligence, and neither was technological assistance in the newsroom. The more useful question isn't whether journalists should use AI. Many already do. The question is how they use it and whether a human being remains responsible for the finished product.

That answer ought to be easy: the journalist and the editor.

Newsrooms have always used technology

There is something amusing about treating technological assistance with research and writing as an entirely new ethical crisis. Newsrooms haven't operated with pencils, Underwoods and rotary telephones for quite some time.

Spell check automatically catches mistakes. Grammar software analyzes sentences. Search engines find information in seconds that once required an afternoon in a library. Electronic court databases save trips to courthouses. Smartphones record interviews, shoot photographs and video, transmit stories and provide immediate access to public records. Software can transcribe an hour-long interview before a reporter finishes lunch.

Journalists don't attach a disclosure saying a story was produced with assistance from Google, Microsoft Word, spell check, an iPhone, Ctrl-F and a calculator. Those are simply tools of the modern trade.

Artificial intelligence is obviously more powerful, particularly because it can generate language and synthesize enormous amounts of information. That creates additional risks, but it also creates opportunities that journalists would be foolish to ignore.

Indianapolis newsrooms are already using it

That's what makes the Capital Chronicle's May 20 examination interesting. Indianapolis journalism isn't standing outside the AI revolution shaking its fist at ChatGPT. Newsrooms are already inside experimenting with it.

The Capital Chronicle reported that IndyStar has AI guidelines, while WFYI uses AI for transcription and captioning. Mirror Indy uses it for transcription, Chalkbeat Indiana has used transcription technology for meetings, and Capital Chronicle itself uses AI transcription while having reporters verify quotations against the recordings. Axios has gone further, using AI for tasks that include organizing documents and analyzing data.

None of this should terrify readers. If software can accurately produce a draft transcript of a two-hour government meeting in minutes, there isn't much civic virtue in making a reporter spend the afternoon typing it manually.

Woodward and Bernstein probably wouldn't have rejected a searchable PDF because Ctrl-F violated the sacred traditions of investigative journalism.

Imagine the conversation:

“Bob, I found the document.”

“Great. Search for Nixon.”

“No. That would be unethical. We have to read all 14,000 pages.”

Technology doesn't eliminate journalism. Used intelligently, it gives journalists more time to do it, to ask the questions, compile the facts and find the old yearbook pictures of say a "Peter Donahue" swept under the rug by the mainstreet media until now.

The real opportunity is research

This is where the AI discussion becomes far more interesting than whether a computer helped polish a paragraph.

Consider a newsroom receiving 5,000 pages through a public-records request. Not long ago, somebody might have spent days going through those pages one by one. Today, technology can help organize the material, locate names, identify dollar amounts, build timelines, compare documents and flag sections that deserve closer examination.

A reporter still needs to read the important original documents and determine whether the machine understood them correctly. But instead of spending eight hours looking for the needle, the reporter can spend more of those eight hours figuring out why the needle matters.

For small and independent news organizations, that could be transformative. Research capabilities that once required teams of reporters, researchers and data specialists are becoming available to publications operating with a fraction of those resources.

That's not the death of investigative journalism. It could produce considerably more of it.

The world's fastest research assistant — who sometimes makes things up

Of course, there is a catch. AI can behave like an incredibly fast research assistant who occasionally doesn't know the answer but is too embarrassed to admit it.

That's why journalists can't outsource judgment.

If AI identifies a court decision, read the decision. If it points toward an important government statistic, find the underlying report. If it says a politician made a particular statement, locate the speech, interview or transcript. If it discovers something suspicious in a contract, examine the contract and start making telephone calls.

AI can tell a reporter where to look. It shouldn't automatically determine what the reporter believes.

That distinction isn't particularly revolutionary. Good journalists have always distinguished between a tip and evidence.

Google already changed the profession

There is also some historical amnesia surrounding this debate. Google transformed reporting long before ChatGPT arrived, along with Perplexity, Grok, Claude, Gemini, and many others, some offsprings of the real AI platforms.

A reporter investigating an Indiana business executive in 1990 might have searched newspaper archives, visited courthouses, called government offices and spent days assembling background information. Today much of that preliminary research can happen before the first cup of coffee gets cold.

Nobody argues that Google has become the reporter because a journalist found a court case through a search engine. The journalist still determines what questions matter, evaluates the information, talks to people and decides what can responsibly be published.

AI is another enormous jump in that process because it can do more than locate information. It can help people work through it.

That makes human judgment more important, not less.

AI isn't just for reporters

The biggest impact may ultimately occur outside newsrooms. Maybe all investigators should and do already use AI.

Government produces mountains of technically public information that relatively few members of the public can realistically digest: budgets, regulations, court opinions, meeting minutes, audits, contracts, environmental reports, campaign-finance filings and hundreds-page pieces of legislation.

Putting a 600-page PDF on a government website satisfies the technical definition of transparency. Whether many taxpayers will read it is another question.

AI can help ordinary people interrogate those documents in ordinary English. A homeowner can ask what an ordinance means for a property. A taxpayer can ask where money in a budget is going. Someone reading a court decision can ask for the legal jargon to be explained. A citizen comparing versions of legislation can identify what changed without needing a law degree and an empty Saturday.

The answers still need to be checked when they matter, but making complicated public information easier to navigate has genuine democratic value.

For generations, access to information often depended upon having the money, education, professional training or institutional resources necessary to find and understand it. AI has the potential to narrow that gap.

Don't confuse typing with reporting

Perhaps the strangest part of the debate is the assumption that journalism primarily consists of writing sentences. Writing is simply the part readers see. Don't confuse reading with thinking, analyzing and understanding.

Reporting happens before that. It is noticing that two government numbers don't match, wondering why a contract went to a particular company, discovering that an official statement contradicts a document, finding someone everybody else forgot to interview or asking the question nobody else thought to ask.

No computer automatically possesses that curiosity.

AI may help search 10,000 documents for the answer, but somebody still has to recognize that there is a question worth asking.

That's the journalist's job. What the powers don't like is they no longer control the discourse and when they search their names they now get to read the real news about how criminal they have been in the rise to the top of the mountain.

Humans have been getting things wrong for centuries

Warnings about AI hallucinations are justified, but let's not romanticize the alternative. Human journalists have printed wrong names, bad statistics, incorrect election results and false information since newspapers were invented. Corrections pages existed long before Silicon Valley.

Artificial intelligence creates new ways to make mistakes, and some are particularly dangerous because fabricated information can look convincing. That means newsrooms need sensible rules about its use, especially involving quotations, confidential information, photographs and important factual claims.

But the solution isn't technological purity. It's editorial responsibility.

A lie doesn't become journalism because somebody manually typed it into Microsoft Word, and a verified fact doesn't become less true because software helped a reporter find it.

A remarkable opportunity for independent journalism

Perhaps the most consequential part of this revolution is what it could do to the economics of journalism.

"HE has been researching writing original articles and google even has recognized the quality and quantity of our staff, team, and reporters news stories, and we arent going to let some rag tell you that AI is what you reading here. We are smart, educated, and serious people who have news as our mission. Real news is needed in Indiana and the Hoosier Enquirer which has been expanding both north and south."

-Donna Kavanaugh, CEO

When HE uses a generated image by AI, it is noted as such. This practice will continue until the industry updates the rules and requirements, which only the Hoosier Enquirer from Indiana submitted a written comment to prior to proposed Journalism Ethical Rule Making began this year.

Large newspapers traditionally possessed advantages that smaller publications simply couldn't match: research departments, archives, specialized reporters, data teams, copy desks and enough employees to spend weeks investigating a complicated subject.

AI doesn't erase those advantages, but it can narrow them.

An independent Indiana newsroom can potentially analyze thousands of pages of records, compare years of government spending, search lengthy court histories and organize complicated datasets with resources that would have been unimaginable for a small publisher not long ago.

That could make established media uncomfortable.

It should make readers excited.

More people capable of examining government records means more people capable of asking government questions. Competition among journalists is healthy, particularly when the competition is over who can uncover something important first.

Keep the computers — and keep the editors

Hoosier Enquirer isn't afraid of artificial intelligence. We're considerably more worried about artificial stupidity.

Use AI to research, organize, transcribe and analyze. Use it to help navigate enormous public records and identify questions worth investigating. Let computers handle repetitive work that once consumed hours of a reporter's day.

Then let journalists do journalism, tell the stories and file corrections and retractions where needed.

Read the documents. Make the calls. Challenge the assumptions. Give people an opportunity to respond. Correct mistakes. Most importantly, put a human being behind the finished product who accepts responsibility for it.

Artificial intelligence doesn't have a reputation to protect, a reader to answer to or an editor who can fire it. A newsroom does.

Journalism isn't defined by whether somebody used a pencil, an Underwood, Microsoft Word, Google or artificial intelligence. It's defined by whether reporters found something worth telling the public, established that it was true and presented it fairly enough. Grow up! And stop your readers throw up!