Journalism’s AI reckoning has arrived: Taking power back from Big Tech

The most successful news organizations have viewed AI not as a tool for content creation, but as a means to create engaging new experiences. But we just can’t avoid repeating the mistakes of the past; instead, we should be truly pushing for real change that would benefit newsrooms nationwide.”

Rob Tornoe, digital editor and writer, The Philadelphia Inquirer

Facebook videos. Google ads. Instant articles. Accelerated mobile pages. Search engine optimization. Now tech companies are touting artificial intelligence (AI) with no intelligence as the next best way to reinvent journalism in the digital world. Plenty of news organizations are taking the bait — hook, line and sinker.

We shouldn’t have believed them then. And we certainly shouldn’t trust them now. Big Tech companies aren’t here to preserve or protect journalism. They’re here to help destroy it in their unrelenting quest to control every eyeball (along with their dollars) every minute they can.

The only difference with AI is that tech companies aren’t trying to hide their goal of not helping news organizations, but of replacing them. Well, everyone except Google, who continues to gaslight journalists that AI Overviews and other algorithm changes will help grow search traffic and offer “relatively stable” click volume.

Any editor with a Chartbeat or Parse.ly dashboard knows that’s not the case. News organizations nationwide, including CNN, NBC News, The Washington Post and Business Insider, have experienced dramatic drops in search traffic over the past year as Google (which essentially has a monopoly on the industry) keeps readers on its own sites, reading AI-powered results. Users can even have Google create computer-powered news hosts to banter and read your town’s top stories, all without sending a single dime or reader your way.

“Google is shifting from being a search engine to an answer engine,” Nicholas Thompson, the CEO of The Atlantic, said in an interview with The Wall Street Journal last year.

Yet despite the outward threat, news companies continue to fall prey to the idea that the solution is to embrace the very technology that’s working to undermine them and the industry. It’s a technology with a well-known tendency to make up facts and information — “hallucinations,” as they’re known in the tech world.

The results are predictable

The most recent embarrassment comes from The Washington Post, where an AI-powered tool that creates personalized podcasts (surprise!) made up facts and expressed misplaced opinions.

It was Post staffers who flagged major errors to Semafor’s Max Tani, including “misattributing or inventing quotes and inserting commentary,” leading to cases where a source’s quotes were conveyed as the paper’s own opinion.

Of course, AI-powered embarrassments are ubiquitous across the news world, hitting organizations such as Wired, Bloomberg, Business Insider and The Associated Press. Even Apple had to pause a news alert feature that offered wildly inaccurate information about BBC News articles. Yet that hasn’t done much to alter the drumbeat toward deploying these tools in front-facing efforts for readers without proper editorial oversight.

Can AI-powered tools be useful? Undoubtedly. At The Philadelphia Inquirer, where I work, we’ve created an advanced search tool that allows readers to comb through the paper’s 195-year archive. At the Norwegian news organization Verdens Gang, AI-powered tools help reporters easily create content featuring their stories that can be shared across multiple platforms. CalMatters created an AI-powered tool called “Digital Democracy” in California that “tracks every word spoken in public hearings, every dollar donated to politicians, every bill introduced, every vote cast and more.”

The most successful news organizations have viewed AI not as a tool for content creation, but as a means to create engaging new experiences. But we just can’t avoid repeating the mistakes of the past; instead, we should be truly pushing for real change that would benefit newsrooms nationwide.

Taking back our power

So, how can news organizations really take back some of the power stolen by Big Tech? Instead of focusing on the shiny new toy that promises to rescue journalism and restore the big profit margins of the print days, there are a number of good ideas on the floor that I think we should all back and push.

One of my favorites was suggested a few years ago by Cory Doctorow, the journalist and activist known for coining the term “enshitification” to describe tech companies making their products simultaneously worse and more expensive. One of Doctorow’s many ideas is to push Congress to pass a comprehensive privacy law that would effectively ban the “behavioral advertising” tech companies use to squeeze billions of dollars out of news organizations.

A ban would mean tech companies would lose their digital advertising monopoly, which is fueled by all our private data. Instead, contextual ads based mostly on a publication’s content would be much more attractive to ad buyers and more lucrative to publishers.

An alternative idea comes from Republican Senator Mike Lee of Utah, who has sponsored legislation that would force Google and Facebook (and likely Amazon also) to divest from their all-powerful digital ad businesses. This would end the ad monopoly these large tech companies have on the digital ad market — and the ruinous rates they’re able to charge publishers.

“Google uses its pervasive market power across the digital advertising ecosystem, and exploits numerous conflicts of interest, to extract monopoly rents and stack the deck in its favor,” Lee wrote in a social media post about his legislation. “These monopoly rents function as a tax — upwards of 40% — on every ad-supported website and every business that advertises online, collectively a huge segment of the modern economy.”

We should also look to Europe, where its recently enacted AI Act requires companies to adhere to several rules governing AI-generated content. For images, that means every photo partially generated by an AI tool must be clearly marked. The same goes for text, unless it’s edited and fact-checked by humans.

That would quickly level the playing field here in the U.S. and help separate AI-generated slop from the trusted content news organizations provide. It would also be another avenue for redressing the balance between tech giants and the journalists who cover them.

Unfortunately, I’m not optimistic, especially for the next three years with Donald Trump in office. Going the legislative route rarely works in journalism’s favor. Just look at California and the debacle involving the California Journalism Preservation Act, which would have required Big Tech platforms to pay or negotiate with news outlets in the state to use their work. It’s not as if Google didn’t have the money. It spent nearly $11 million lobbying in California to kill the legislation in 2024, according to CalMatters.

Instead, Google agreed to spend $55 million over five years to grow a fund for local newsrooms and $70 million for… an AI “accelerator” program. Guess who ultimately benefits?

At this point, the best I can say is journalists aren’t the only suckers. Disney just signed a deal with OpenAI that will see Mickey Mouse and the gang legally exploited by AI-slop videos, flooding our social media feeds and ultimately damaging their brand.

I guess misery really does love company.

Rob Tornoe is a cartoonist and columnist for Editor and Publisher, where he writes about trends in digital media. He is also a digital editor and writer for The Philadelphia Inquirer. You can reach him at robtornoe@gmail.com.