ProPublica journalists walked off the job for 24 hours, after more than two years of negotiations that failed to yield a deal for a union contract that would have included terms around AI and a ban on AI-related layoffs.

Meanwhile, in Italy, the country’s main journalists’ union called for two strike days over publishers refusing to accept basic rules on the use of artificial intelligence. And at The New York Times, according to Axios, editorial union leaders told the newspaper’s management its AI standards are too vague and inadequate, creating editorial problems and trust issues.

As AI is becoming a defining issue for labor unions, I spoke with four journalism union representatives from the United States, the Philippines, and Greece to find out how their organizations are protecting their members from any potential labor changes that AI might bring.

The unions versus AI

No union I spoke to reported having any of their members being replaced by AI. But one of their central concerns has always been ensuring human staff is protected as these technologies become widespread. Collective bargaining agreements help enact these protections. Some agreements implicitly state that AI cannot be used to displace a member of the staff, like the News Media Guild, while others mandate higher severance pay if layoffs are AI-related, such as the PEN Guild.

However, AI use at work raises many complex issues beyond layoffs, said Tony Winton, chief administrative officer of the News Media Guild, which represents newsrooms like the Associated Press and The Guardian in the United States. Unions have the right to bargain not just over whether jobs remain, but also over working conditions and how AI changes the way people do their jobs.

“The more difficult issue is which uses are allowed, short of something that actually changes the size of the workforce, and there are a lot of very thorny issues here,” he said. “We have an active working group of members who want to expand this conversation with the AP. The contract language we have is good. But as more and more uses are being found for the technology, we need to have a conversation.”

The specific uses of AI in a newsroom, and how they impact the work of journalists beyond layoffs, is something that all union representatives I spoke to cited as grievances they have brought to their management. While they care about core issues like jobs, pay, and working conditions, Winton said, AI also raises serious concerns about journalistic accuracy, for example, that managers need to address.

“AI has struggled with a lot of fabrication problems,” he said. “So, for a person with a byline and a public identity, AI is a real concern. You don’t want to incorporate inaccurate work into your reporting that affects not just journalism quality, but also the reputation of the person whose name is attached to the story itself.”

Ariel Wittenberg is a public health reporter and the unit chair of PEN Guild, which represents workers at Politico and E&E News. Like Winton’s, her union has not seen layoffs due to AI yet, but her concerns extend towards the way AI is used, and how it can impact journalists’ work and journalism ethics.

She described two recent incidents at Politico, where managers were required by contract to warn the union in advance and negotiate before using AI in ways that meaningfully affect employees’ job duties. Politico ignored this clause and deployed two AI initiatives without telling them: one used AI to generate written coverage of the Democratic National Convention and the other one was a deal with Capital AI to automatically produce reports.

“We think they violated the contract, which says that any AI use has to be done in accordance with Politico’s standards of journalism ethics and with human oversight,” said Wittenberg. “If something is coming back with inaccuracies, if it’s not following our stylebook in other ways, and there are no corrections policy applied, that is not up to our political ethics.”

“An existential threat”

Establishing protections on AI-related issues hasn’t been easy for journalists working in other latitudes. A newsroom manager who is also a director for the National Union of Journalists of the Philippines (NUJP) spoke to me on the condition of anonymity about how difficult it is to establish protections for workers on these kinds of issues.

He said most newsrooms just have general provisions of using AI responsibly and ethically. But nowhere is it stated that AI will not be used to replace journalists. “This is an existential threat,” he told me. “My hope is that at some point [managers] will realize it and then we will have to adjust our policies on it.”

The Philippines’ national union does advocacy work, whereas specific employer unions are the ones with bargaining power. While there is no authoritative count, the latter ones in the Philippines are limited in number, unevenly distributed, and much less institutionalized than in European countries. The newsroom of the manager I spoke to, for example, doesn’t have a union in place.

“[The] most the NUJP can do is to issue statements to create noise, to try to advance the conversation, and to call attention to certain issues,” he said. “The most you can do is to recommend. We have to set these policies in stone and encourage media owners to craft a policy that would protect their workers from the threat of AI.”

Journalists in other countries face a similar challenge. Greek journalist Sotiris Triantafyllou, president of the Panhellenic Federation of Journalists’ Union, describes AI adoption in his home country as not quite as expansive as in Northern Europe. This has allowed his union to be ahead of the curve domestically. In 2025, for example, they launched a code of ethics now adopted by the five unions of the federation.

“Now we are in discussions with managers and media owners. I don’t know what will happen in the future. But for now they agree with us, and I think they are in a mood to protect journalists,” Triantafyllou said.

“Like nailing Jell-O to a wall”

What all the union representatives I spoke to are looking for is a bottom-line commitment to human-led journalism and that AI does not take over skilled labor. There is broad support in using AI “housekeeping” tasks like transcription, translation, and summarization of large datasets. But pushback arises when managers implement tools that automate creative and journalistic work.

“We try to protect the central role journalists play because we believe that AI cannot replace them,” Triantafyllou said.

Unions often have to play whack-a-mole to deal with all the potential effects AI can have on workers. The initial question was perhaps “Will AI come for my job?” But now a myriad of other questions arise: if an employer sells journalistic information to a model, should employees who produced that content be compensated? Is the use of AI optional or will employees be replaced if they don’t adopt it? Will there be universal training for employees to apply these tools?

“It’s a moving target. It’s like nailing Jell-O to the wall, because you think you’ve got something done, and then the technology changes again,” Winton said. “When you are hired to do a job, you are hired to write a story for a publication, not to be part of this blob of AI that goes on forever. There’s a lot of interesting things that people are thinking through.”

Some news organizations, for example, are now trying to increase their output with the help of AI and AI-assisted reporters, such as U.K. local news publisher Mediahuis. Recently, Fortune editor Nick Lichtenberg came under scrutiny after a profile detailed how he used AI to crank out more than 600 stories. On these use cases, the journalists I spoke to believe in having a seat at the table: as AI-writing is becoming an unavoidable reality of journalism, journalists should have a say in how AI is used in their newsrooms rather than just executives looking to adopt the technology.

The newsroom manager and union director from the Philippines believes that while AI writing in journalism is seen as deeply unsettling because it threatens human creativity, authenticity, and editorial craft, its spread is still inevitable as economic pressure will push newsrooms to adopt it.

“It’s sad and tragic in a lot of ways, and many of us are mourning the kind of journalism we are used to, but the reality is ChatGPT, Gemini, and others are already capable of replicating the way humans speak and write, and they’ve been able to do so for quite some time now,” he said.

Despite these ongoing challenges, unions seem more important than ever. The representatives I spoke to highlighted a number of victories, from proactive negotiation with management in the case of Triantafyllou in Greece to providing binding arbitration in the United States.

“AI is something that is already impacting our industry, and union contracts are one way that journalists can have a say in how AI is deployed, rather than leaving those decisions up to news executives or corporations,” said Wittenberg.

A difficult balance

Few industries show financial distress as clearly as the news industry: repeated waves of job cuts, declining engagement, and precarious business models. In light of these existential challenges, AI has been presented as both a problem and an opportunity for growth.

No newsroom wants to be left behind, and some are resembling Silicon Valley in their language of adoption, pursuing rapid experimentation, “content scaling,” and liquid content.

Wittenberg has found AI to be a useful tool in handling large data sets or doing menial tasks like transcriptions. But she thinks some newsrooms have lost sight of why audiences come to them: because they want accurate and factual news.

“In the rush to innovate, news organizations think they are competing with tech companies,” she said. “The reality is that we are still news organizations and that means that we have an obligation to our ethics and to give our readers accurate factual news and to be held accountable when we make mistakes.”

My source in the Philippines admitted that protecting media workers from AI’s potential harms will be difficult because the news industry largely regulates itself and media owners are not naturally incentivized to put strong protections in place.

“They are looking at how they can make news more efficient, how they can save more money, how many employees they can let go because AI can do the work that they’re doing,” he said.

In his view, despite having limited power, journalists and their unions should still push to protect their own rights and the industry as a whole as many concessions will happen due to public pressure and broader public opinion.

“There’s always been that kind of divide between those who own the news media and those who are the news media,” he said. “As journalists, we have to be prepared because this is going to be an uphill battle.”

In 2018, Joseph Poliszuk fled Venezuela. That year, after exposing corruption in then-President Nicolas Maduro’s administration, he had become the target of lawsuits by wealthy Maduro loyalists. He and several of his colleagues at the independent outlet Armando.info packed up their lives and fled the country under threat of imprisonment.

For years, Poliszuk had published stories on Southern Venezuela, which is made up of sparsely populated states that cover large swaths of the Amazon Basin and the Orinoco River Basin. Through field reporting, Poliszuk had exposed illegal gold mines, narcotrafficking operations, and crimes against indigenous groups scattered throughout the region’s rainforests. Now in exile — first working from Colombia, then Mexico — Poliszuk was forced to reimagine how to do his work from thousands of miles away. He began experimenting with satellite-based investigations.

Satellite imagery has long helped investigative journalists gather intelligence on conflict zones and track changes in remote landscapes. Now, in a new wave of satellite-based investigations, reporters are leaning on machine learning models to automate parts of this work and scale up their analysis to an unprecedented degree.

This innovation is most visible in environmental journalism. Poliszuk is just one in a cohort of South American investigative reporters who have used geospatial data and AI-powered pattern recognition to track illegal mining, large-scale logging operations, and cattle ranching across the Amazon.

As illegal gold mining spiked during the COVID-19 pandemic, Poliszuk knew there was a story in documenting the growth of these mines across Venezuela’s rainforests. But manually combing through the satellite images for over 50 million hectares of rainforest wasn’t practical. Poliszuk wondered if he could train a machine learning model to detect the scars of mining pits in these images, as well as the neighboring airstrips that are cut into dense vegetation and used to transport minerals.

With financial and editorial support from the Pulitzer Center’s first Rainforest Investigations Network (RIN) fellowship and technical support from the nonprofit Earth Genome, Poliszuk was able to do just that. In January 2022, he co-published his first article using the custom machine learning model in a series in El Pais titled “Corredor Furtivo [Clandestine Corridor].”

Poliszuk was able to identify 3,718 gold mining locations in the Venezuelan states of Amazonas and Bolívar. Some of those mines were operating inside protected indigenous lands and Canaima National Park, which is home to Angel Falls, the world’s tallest waterfall. By crosschecking maps identifying mining activity with crime data from Venezuelan authorities, Poliszuk was also able to determine whether the mines were run by Venezuelan syndicates, Colombian guerilla groups, or Brazilian garimpeiro (prospectors).

The week after Poliszuk published one of his first stories in the El Pais series, the Venezuelan military announced that it had bombed several illegal airstrips operating in the region.

“I have 20 years’ experience covering [illegal mining]…thanks to this technology I can show people the dimension of this phenomenon,” Poliszuk told me. “Thanks to this movement, we have understood that we can track by the air what we cannot prove on foot.”

The view from above

Even for journalists who aren’t working in exile, field reporting in the Amazon comes with a litany of accessibility issues and security risks. Poliszuk said a trip from the Venezuelan capital to one of the mines in the state of Amazonas involves a two-hour flight, a six-hour car ride, a four-hour boat ride, then another four-hour trek through the jungle — often through dangerous territory occupied by armed militia. These same groups often hold monopolies on oil and gas in the region, which can make fuel expensive and difficult to procure.

“It’s dangerous. It’s challenging. You cannot go there like you can go from Boston to Washington, or Caracas to Maracaibo,” he said.

The same year Poliszuk pitched his project to the Pulitzer Center, Brazilian journalist Hyury Potter incubated a similar investigation with the RIN fellowship. He also used machine models from Earth Genome, which collaborates with many journalists to conduct AI-based environmental and human rights investigations. Potter went on to publish several major investigations in Intercept Brasil that identified hundreds of previously unreported airstrips in the Brazilian Amazon and documented the explosion of illegal gold mining on protected indigenous lands. The New York Times published its own investigation based on the satellite imagery analysis, collaborating with Potter and the Pulitzer Center in the reporting process.

“It was like a think tank trying to figure out how to do this work,” said Poliszuk of his time in the RIN fellowship. “It was a very good time to think about a new journalism — another way of doing things.”

Based on the strength of these early investigations, the Pulitzer Center decided to build a dedicated platform that uses machine learning to track mining activity across the nine countries that are part of the Amazon Basin. Earth Genome built the interface and contributed the underlying geospatial detection models. The nonprofit advocacy group Amazon Conservation contributed fundraising support and helped develop impact metrics. In 2022, the three partner organizations launched Amazon Mining Watch.

“That was the beginning — inspired by the works of Joseph and Hyury, we were able to extrapolate and cover the entire Amazon,” said Gustavo Faleiros, the former director of environmental investigations for the Pulitzer Center.

The earliest days of Amazon Mining Watch relied on small, task-specific machine learning models. These models were trained by Earth Genome itself and customized only to identify gold mining sites and airfields in satellite imagery. These days, though, Earth Genome is experimenting with more powerful geospatial foundation models — models pre-trained on huge amounts of data, including satellite imagery, but also radar, land cover, and elevation data.

It’s likely these larger models will make geospatial investigations even more accessible to journalists, and not just ones covering the Amazon or illegal mining.

“In the same way that people figured out how to do unsupervised training of AI models for text — techniques that grew into large-language models — they have done the same thing in the geospatial data space,” said Edward Boyda, a physicist and co-founder of Earth Genome. “With very little additional input from a user — maybe just a few examples — the model can be effectively tuned to detect a wide variety of objects on the Earth’s surface.”

Beyond the Amazon

The Pulitzer Center and Earth Genome are now partnering with the nonprofit Code for Africa to bring a similar platform to the African continent. Earlier this month, the organizations announced the launch of Africa Mining Watch. The platform will use geospatial detection models to track mining operations across the tropical bend, a region that includes the Congo Basin, the world’s second-largest tropical rainforest. It’s expected to launch publicly in July.

On Earth Day last week, 25 journalists from across Africa took part in a seven-hour virtual mapathon to experiment with the platform and test out its ability to identify mines in their coverage areas.

“My hope is that Africa Mining Watch will be a platform that’s not as connected with the gold mining story, but with the strategic minerals story,” Faleiros said, pointing to the cobalt, copper, and coltan mines found across the Congo Basin.

Earth Genome is also building its own platform to harness these foundation models for journalists. The tool, Earth Index, allows a reporter, researcher, or policy maker to go into the platform and select a region on the world map. After they select examples of the thing they are interested in identifying — say, an artisanal gold mine in Ghana — the platform highlights other potential gold mines in the region.¹

In its invitation-only beta stage, Earth Index has been used to investigate illegal logging in Albania, commercial flower farms in Uganda, and palm oil production in Brazil. Boyda says they plan to release Earth Index publicly in late April.

“The idea with Earth Index is, instead of giving people the data, give them the tool to make their own data,” said Boyda. “Somebody who’s working in a specific area will know that context better than we ever could. With this tool, they can go and build the data set that they want.”

This story has been updated to correct where Earth Genome is based.

1. While LLMs have been criticized for their environmental footprint, including the energy and water consumption of data centers used to train them, Boyda says the models underlying Earth Index are a fraction of that size. Currently, Earth Index uses a foundation model built by the Technical University of Munich, which has about 20 million parameters, as opposed to the trillions likely found in the latest commercial LLMs. Processing two years of global embedding for the latest Earth Index release used 190 kWh of electricity, which comes out to about a week of an average house’s electricity use, according to Boyda.