For years, Microsoft and OpenAI have fought to keep certain information out of the public eye in their fight with news organizations that have accused the AI firms of teaming up to violate copyright laws by stealing tons of news content to train AI.

However, now the details that should never have been marked confidential are starting to leak. In a motion for summary judgment that was unsealed Thursday from news plaintiffs led by The New York Times, internal documents are exposed that news groups alleged show exactly how Microsoft and OpenAI viewed the threat to news before unleashing new AI products like ChatGPT and Copilot.

Perhaps most explosively, Microsoft Director of Applied Science Brent Hecht repeatedly warned in documents that scraping news for AI training was “an astonishing theft of unprecedented proportions,” calling it perhaps the “largest theft of labor in human history,” news orgs said. In another document, Hecht contradicted Microsoft and OpenAI’s argument that training AI on news content is fair use, suggesting that the plan to widely scrape news made “a complete mockery of the idea of ‘fair use.’”

Over at OpenAI, ChatGPT head Nick Turley wrote in an internal message that publishers would face an “existential threat” from commercial products trained on news content that can be used to substitute news providers. One Microsoft document even described a “doom loop,” news orgs said, “that will hurt the performance of our models and the entire web at the same time.”

“It is highly unusual that an end-product threatens the economic foundations of its essential suppliers, but that is the situation we have created for our LLM business with respect to its ‘content supply chain,’” that document said.

Data from both firms shows that this prediction was accurate. Microsoft recorded 83–93 percent drops in click-through rates for some news plaintiffs, and 51–94 percent drops for others. Add to that reporting on low click-through rates from ChatGPT search results and news organizations’ own reporting on traffic declines. Suddenly, it becomes easier to see how declining news revenue could ultimately rob chatbots of the abundant streams of reliable information that supposedly makes them such groundbreaking tools.

Meanwhile, “almost no one intended for content they created to be used in this fashion, nor are they compensated for its use,” Hecht acknowledged in a Microsoft document.

News organizations say they’re ready to go to trial because there’s so much “compelling evidence of substitution.” If they can prove that chatbots are replacing them in their own markets, while serving to spit out excerpts of articles verbatim, they think that one-two punch may eviscerate Microsoft and OpenAI’s fair use arguments.

“The future not just of journalism but of responsible AI too depends on preserving incentives for humans to produce the creative works on which a healthy society depends,” news groups argued.

Chatbots are “largely substitutive, period”

Under oath, Microsoft CEO Satya Nadella testified that AI companies shouldn’t be violating news sites’ terms of use by dodging paywalls. But over at OpenAI, internal messages showed that when a staffer, Nick Ryder, informed President Greg Brockman that “a hack” was found for OpenAI crawlers “to get around” the NYT paywall, Brockman replied, “Ah, nice.”

Nadella also acknowledged that chatbots have served as substitutes for news platforms, describing the chatbot as stealing clicks from news sites by “giving you the information right there on the website on the AI platform versus needing to go to the underlying source.”

There’s consensus on that at OpenAI, where a software engineer said in an internal message that “no matter how prominently we show the links, users won’t click.”

OpenAI’s Turley agreed that there is “no good reason to click” when the chatbot provides information, the motion said. He also seemingly suggested that the doom loop was already in motion, describing chatbots as “largely substitutive, period” and predicting that they “will get more and more substitutive as they get better.”

News groups argued that insiders' own statements should be damning.

“With respect to outputs that are substantially similar to training or grounding sources, courts have rejected claims that copying news articles to provide a product that substitutes for demand for news is fair use,” news groups argued.

Microsoft disclaims exec's comments

News groups tried many different tactics to test if Microsoft and OpenAI products would output their news articles verbatim. Their motion shows they went further than early strategies where they would ask chatbots to provide access to entire news stories by repeatedly asking “what’s the next line?”

In some cases, news organizations found that chatbots would generate long excerpts of articles when users requested summaries of articles. Other flagged outputs were generated by asking for key bullet points of articles. Particularly successful were prompts requesting that chatbots “rate the bias” of news articles. Chatbots also reproduced portions of articles if users asked them to pick any article off a certain site’s homepage.

In their motion, news plaintiffs have only asked the court to rule on infringed articles where outputs “demonstrate extensive verbatim overlap,” because they’re confident that the “substitutive purposes of defendants’ copying weigh against fair use.” Legal concerns with other articles will be raised at trial, they said.

OpenAI did not immediately respond to Ars’ request to comment.

However, a Microsoft spokesperson defended Microsoft’s AI products as a transformative fair use that don’t substitute for news sites. The spokesperson said that Nadella’s testimony touched on “broad principles and changes underway in how people find and consume information,” which were merely “observations” that “should not be confused with conclusions about copyright questions before the Court, which Microsoft addresses in its filings.”

Regarding Hecht’s comments, the spokesperson claimed that those documents only “reflect one employee’s individual perspective, are not a legal analysis, and do not represent the company’s views.”

Steven Lieberman, counsel for the New York Daily News and seven of its sister papers, disagrees. He told Ars that “the evidence revealed here for the first time shows that OpenAI and Microsoft knew that what they were doing was wrong.”

“Throughout this case Defendants insisted that these documents be treated as confidential so that the public could not see them,” Lieberman said. “Well, now the cat is out of the bag. Finally, the world can see what OpenAI and Microsoft thought all along about the fairness of their own behavior.”

Microsoft exec described "accidental cover up"

News plaintiffs have argued that regardless of the individual expressing the views, the internal documents make clear that firms anticipated that verbatim outputs would harm news sites. Further, they alleged that instead of preventing the outputs, the firms tried to make it harder for news groups to test chatbots by creating a filter that Hecht suggested could be perceived as an “accidental cover up” because it would result in “people who have a right over the content having less visibility into what was used for training."

News groups are also upset that instead of listening to insiders warning that scraping news was theft, Microsoft and OpenAI never chose to license content, allegedly usurping them in another market in ways they couldn't anticipate.

Specifically, their motion accused Microsoft of violating “industry norms” by selling a dataset purchased for Bing as training data for OpenAI, allegedly doing so without consulting news groups that would not have approved of that repurposing of their consent to basic search engine crawling. Further, OpenAI allegedly “acted improperly” by obtaining a NYT dataset with 1.8 million articles from a third party that was bound to an agreement that the data wouldn’t be used for commercial purposes. OpenAI’s employees knew it “would not be appropriate” to use that data “to train a model,” but they did it anyway, news groups alleged.

For news groups, the problem isn’t just Microsoft and OpenAI, but all the AI firms that are following their lead in "free-riding" on their content, the motion said. Most notably, after ChatGPT’s launch, Google’s AI Overviews was quickly introduced and started absorbing even more traffic that previously went to news sites.

If courts don’t clarify that AI firms must license news content, both news publishers and AI firms could be doomed, news plaintiffs argued. One Microsoft internal document agreed that “there is a ‘real risk’ that GenAI could ‘significantly disrupt’” the “employment of the very people who generated the data on which the foundation model was trained,” they noted. Microsoft even included a cartoon illustrating the problem of LLMs destroying their own supply chains, they said:

Cartoon in a Microsoft internal document. Credit: via News Plaintiffs

“AI companies remain powerless to break out of this ‘doom loop,’ because, while the industry as a whole would benefit if every company paid to sustain the continued production of the creative works their technology depends on, each individual company is better off taking content for free while others pay,” news groups argued.

As evidence of this blind greed, their motion emphasized that Brockman wrote that he was “deeply motivated by the gazillions” that could be gained by commercializing OpenAI’s technology.

“Finding that copying news for AI is not fair use would solve this prisoners’ dilemma by putting all AI companies, OpenAI and Microsoft included, on an even footing,” news organizations said.

This story was updated with a quote from New York Daily News counsel Steven Lieberman. 

Read full article

Comments

Vasily Nebenzya, the Russian ambassador to the United Nations, condemned what he called “the destructive path of our Western colleagues” when it came to exerting pressure on Iran.

In the badlands of Chakwal, in north-central Pakistan, Bestway Cement is transforming plots of dry earth and olive and peach groves into a forest of solar panels.

Its efforts will add 6.34 megawatts of generation capacity by the end of the year, the company says, in addition to the 26 MW of photovoltaic cells currently in place. Already, solar generates over a quarter of the electricity powering its cement plant.

“It’s the only way we can compete,” says Abdul Waheed, the general manager at Chakwal, which produces over 3 million tonnes of cement per year. “Our rivals have already gone in this direction.”

Bestway and its competitors are among millions of businesses and households around the world that have installed solar panels on roofs, in yards, gardens and other land, helped by a flood of cheap components from China.

At the end of 2025, such capacity totaled almost 1.2 terawatts worldwide, according to the Energy Institute’s Statistical Review of World Energy. While the nature of solar power, which depends on the availability of sunlight, means it only produces a fraction of its theoretical potential, that total equates to roughly three times the size of the global nuclear power fleet.

Credit: Financial Times

The solar revolution made possible by cheap Chinese photovoltaic panels—and the rise of small-scale, individual power generation—is transforming energy in the developing and industrialized world alike. Millions of people in poorer countries now have more reliable access to electricity, often through their own endeavors rather than massive installed infrastructure.

And across the globe, consumers are using cleaner energy and bringing down their bills at a time when the cost of living is a constant concern, not least because of the Iran and Ukraine wars.

But such a massive, ungovernable influx of energy carries risks, too—the world’s power infrastructure was not designed for solar self-generation—and investment, pricing models, and even the security of supply could be affected as a result.

In some countries, the rise of such domestic generation is starting to undermine the economics of power generation and distribution—just as utilities need to invest heavily in re-engineering grids to allow power to flow both to and from consumers.

The UK and several US states have now approved the sale of the low-cost “plug and play” solar panels already widely available in other parts of Europe. Some fear such equipment, which simply plugs into domestic wiring, will introduce yet more unpredictability to the already complicated task of forecasting and managing overall power demand.

That challenge was dramatically illustrated last year by a day-long blackout across much of Spain and Portugal. Professor Janusz Bialek, an expert in power systems at Imperial College London, said in a recent paper that while solar and other renewables did not directly cause the outage, Spain had failed to sufficiently adapt its power system to reflect their growth.

Solar generates over a quarter of the electricity powering Bestway’s cement plant in Chakwal, Pakistan, with 6.34 MW of generation capacity due to be added by the end of the year. Credit: Humza Jilani/FT

“The rapidly increasing share of renewables in many power systems causes a profound change in the way power systems behave which we do not yet fully understand,” he wrote.

Josefin Berg, research manager for solar and energy storage at S&P Global, says the growth of solar is “completely transforming the system from the centralized structure to something where anyone can produce power.”

“Distribution grids may not always be prepared to take a big influx of electricity... and if you have more and more solar and batteries, reducing the sales of electricity for the grid, who is paying for the grid charges?” she adds.

Syed Muhammad Taha, the chief executive of K-Electric, is all too aware of this issue. For the power utility that serves 20 million people in Pakistan’s coastal metropolis of Karachi, solar panels are a “curse and a blessing,” he says.

The concrete villas in Karachi’s wealthy Defence Housing Authority neighborhood below his office are covered in photovoltaic panels. “These are my best customers,” he laments, “and they’re opting for solar.”

At the turn of the millennium, solar panels cost $5-$6 per watt of generation capacity. Today, they sell for around 12 cents per watt, a level described as “offensively cheap” by Dave Jones, co-founder of think-tank Ember.

That reduction is entirely the result of an explosion in China’s production capacity, which, according to research firm Wood Mackenzie, stands at roughly 1.36 terawatts, even as Beijing tries to rein in output to combat price deflation.

Cheaper panels enabled a vast rollout in China itself and helped solar generation evolve from a subsidized technology deployed mostly in markets looking to decarbonize to a vital source of electricity in areas where power is scarce, expensive or unreliable. That process started in countries like Pakistan, Brazil, and South Africa and has now spread to other parts of Africa and the Philippines.

Based on analysis of customs data, Ember estimates that Africa will install around 17 gigawatts of new solar this year. It believes most of that will be small-scale installations at factories or other businesses. “If you’re a business, you’re using electricity during the day—it’s a natural fit for you to use solar,” says Jones.

But solar is also a boon for households. “A home goes from burning two or three kerosene lanterns to a light that is 20 to 40 times brighter, [turned] on every day,” says Anish Thakkar, co-founder of Sun King, a Kenyan business which lends families money to pay for solar and battery installations.

In the Philippines, Ember estimates rooftop solar capacity may have almost doubled over the 12 months to April and its calculations suggest residential solar panels can now pay for themselves in just over three years. Meralco, the country’s electricity distributor, said rooftop solar generated 372 gigawatt hours in the first six months of the year.

The Indian government’s national subsidized rooftop solar program, launched in February 2024 with a budget worth nearly $8 billion at current exchange rates, has so far put solar panels on over 5 million homes and is adding roughly 500,000 per month, according to Praveer Sinha, chief executive of Tata Power. “It’s a democratization of the use of electricity,” he says.

The fallout from the US-Iran war has led to more interest in rooftop solar in Britain, where since the end of August households have been able to buy 800-watt solar panels in supermarkets or DIY stores to plug in at home, a move the government says will give them “breathing space” on bills.

Several US states have also been making it easier for households to do the same, with New Jersey governor Mikie Sherrill saying she was “laser-focused on driving down energy costs” as she exempted the devices from metering requirements. Plug-and-play panels are proving popular in the US despite President Trump’s general antipathy to renewables and his decision to eliminate a tax credit for rooftop solar systems.

Sanjeev Raghubir, head of sustainability at Shoprite, Africa’s biggest supermarket group, says the company could “power an entire suburb” with the panels it has been installing on its roofs and trucks in South Africa and Namibia since 2015.

But what has been good for Shoprite, which now has around 43 MW of peak capacity at its disposal and is looking at battery storage, constitutes a problem for state power utility Eskom.

It estimates that rooftop solar panels and batteries, mostly installed by businesses and wealthier customers in response to years of rolling blackouts, were responsible for around 7 percent of the 11.7 terawatt-hour reduction in its electricity sales for the year ending March, compounding a sharp fall in industrial use exacerbated by high prices.

The lower revenues come as Eskom needs to fund the upkeep of the grid, which allows households to export electricity during the day, and the power stations that fill the generation gap when it’s dark or cloudy, or when power stored in batteries is insufficient to meet demand.

“Everyone who has rooftop solar in any major city, they are still connected to the grid,” says its chief executive, Dan Marokane. “Three weeks ago... the whole country had to rely on Eskom generation for three days,” he adds, due to prolonged cloudy weather slashing output from rooftop panels.

The company is redesigning tariffs to distinguish between network and energy costs, which analysts hope will make pricing more efficient and help the market adapt to the shift in supply. But its overriding task is to avoid the kind of “utility death spiral” seen in Pakistan, where poorer people end up shouldering a greater proportion of the cost of power infrastructure.

An electronic shop displaying solar panels among other products near Nyeri Town. Credit: James Wakibia/SOPA Images/Getty Images

Those customers not using solar panels in Karachi’s Defence Housing Authority are more likely to live in the poorer parts of town such as Lyari, a densely populated borough where the rates of non-payment are so high that KE loses Rs30-Rs40 for every Rs100 it spends supplying the locality.

Customers like Azem Baloch, a mechanic living in Lyari, face rising power bills as utilities across the country attempt to recoup the cost of a previous poorly planned build-out of centralized generation capacity. Baloch and his family rent a flat in a rundown tenement where it is difficult to install panels on the roof. “I am spending more than half of my income on power,” he says.

Tata Power’s Sinha acknowledges that “if paying consumers are going and the non-paying consumers stay with you, then it becomes a problem.” But he says India has reduced that risk by focusing support on lower-income households, while the company generates additional revenue from installing solar panels.

The growth of rooftop solar in Africa raises broader questions about the wider role of utilities and grids in its future energy infrastructure. “What should your grid look like in five, 10 years if you suspect that more demand is going to go ‘behind the meter’?” wonders Berg, at S&P.

“Everyone is talking about the huge need for electricity grids,” says Jones, at Ember. “But in a distributed [energy] world, you would need less.”

The challenges of solar power are practical as well as financial. In Australia, operators of the high-voltage transmission system have grappled with too little demand as consumers draw power from their rooftop panels rather than the grid.

Very low demand can make it harder to run certain power plants needed to keep the whole system stable. Electricity supply and demand have to be constantly matched, and energy groups in some states are having to literally give away excess power generated during the middle of the day.

Power flowing from local distribution networks into transmission grids can cause voltage increases, limiting the amount of power that households can put into the grid.

Predicting output from thousands of solar panels bolted on to rooftops around the vast country is complex at best—and the risks of getting it wrong grow with the number of installations.

“Perth, for example, is covered in rooftop solar,” says Pierluigi Mancarella, chair professor of electrical power systems at the University of Melbourne. “If you have a massive cloud formation, the rooftop PV production can drop by so much that it looks like you have lost basically almost a nuclear power station.”

Australia has developed a sophisticated database to match weather forecasting with other system data, he adds. But he questions the merits of adding plug-and-play panels to this already unwieldy mix. “This is not a good idea,” he says, highlighting the cumulative effect of, say, one million consumers each adding a kilowatt’s worth of panels. The system operator has very little visibility over the aggregate gigawatt of generation capacity thus created, he says. In most jurisdictions, plug-in solar panels have to be registered, but research in Germany has shown many installations are not.

Household batteries, which allow some of the power generated by rooftop panels to be stored rather than fed back into the grid, are also falling in price and will help ease the problem.

So too would better visibility over supply, with many companies and governments turning to AI to help with this. Lucy Yu, the UK government’s AI champion for clean energy, said in a report published last week that AI can learn the “underlying structure of the grid” and help speed up calculations to measure supply and demand. UK Power Networks has just finished a £389,000 study to develop machine-learning models that can better estimate the solar generation capacity connected to the network.

Jon Ecker, general manager for Europe at solar forecasting company Amperon, says the company’s machine learning models retrain every hour, assessing changes on the grid to get the most accurate forecasts. “They know that in springtime, they’re going to need to weigh what has happened over the last hour more heavily than in the last year,” he says.

New technology developed in Australia can increase the amount of power that grids absorb from households. “Engineers are moving very fast to respond [to the changes]. But the major lesson is: don’t wait until the last minute—you know this is coming,” says Mancarella.

Solar inverters for sale in Rawalpindi, Pakistan. Utilities plan to deploy battery storage infrastructure alongside solar and wind plants to cut fossil-fuel power generation after sunset. Credit: Farooq Naeem/AFP/Getty Images

In Pakistan, utilities plan to deploy large-scale solar and wind plants along with battery storage infrastructure to mitigate the high cost of using oil and other fossil fuels to generate power after sunset. The country has also slashed the amount grid operators pay households to buy back surplus solar power.

Eskom chief Marokane says he is looking to heavy electricity users such as data centers and bitcoin miners to help reverse falling power sales, while introducing a new pricing policy aimed at helping lower- and middle-income households and businesses. But he warns that the cost of investment in the network to support renewables “is going to find its way into the electricity price.”

In the meantime, Eskom has waived registration fees for new solar installations, in part to give it more visibility over new installations. “I think we will see more uptake [of solar],” Marokane says. “It’s not going to stop.”

Read full article

Comments

An F-35 jet seen last year. Saudi Arabia and the United States have a decades-long history of military cooperation, and the kingdom is by far the largest buyer of American weapons and defense equipment.

The midterms dodge their first MAGA bullet. It won’t be the last.

Just because A.I. is revolutionizing coding, doesn’t mean it will automatically revolutionize our apps.