When Frank Pine searched Google for a link to a news article two months ago, he encountered paragraphs generated by artificial intelligence about the topic at the top of his results. To see what he wanted, he had to scroll past them.

That experience annoyed Mr. Pine, the executive editor of Media News Group and Tribune Publishing, which own 68 daily newspapers across the country. Now, those paragraphs scare him.

In May, Google announced that the A.I.-generated summaries, which compile content from news sites and blogs on the topic being searched, would be made available to everyone in the United States. And that change has Mr. Pine and many other publishing executives worried that the paragraphs pose a big danger to their brittle business model, by sharply reducing the amount of traffic to their sites from Google.

“It potentially chokes off the original creators of the content,” Mr. Pine said. The feature, AI Overviews, felt like another step toward generative A.I. replacing “the publications that they have cannibalized,” he added.

Media executives said in interviews that Google had left them in a vexing position. They want their sites listed in Google’s search results, which for some outlets can generate more than half of their traffic. But doing that means Google can use their content in AI Overviews summaries.

Publishers could also try to protect their content from Google by forbidding its web crawler from sharing any content snippets from their sites. But then their links would show up without any description, making people less likely to click.

Another alternative — refusing to be indexed by Google, and not appearing on its search engine at all — could be fatal to their business, they said.

“We can’t do that, at least for now,” said Renn Turiano, the head of product at Gannett, the country’s largest newspaper publisher.

Yet AI Overviews, he said, “is greatly detrimental to everyone apart from Google, but especially to consumers, smaller publishers and businesses large and small that use search results.”

Google said its search engine continued to send billions of visits to websites, providing value to publishers. The company has also said it has not showcased its A.I. summaries when it was clear that users were looking for news on current events.

Liz Reid, Google’s vice president of search, said in an interview before the introduction of AI Overviews that there were hopeful signs for publishers during testing.

“We do continue to see that people often do click on the links in AI Overviews and explore,” she said. “A website that appears in the AI Overview actually gets more traffic” than one with just a traditional blue link.

On Thursday afternoon, Ms. Reid wrote in a blog post that Google would limit AI Overviews to a smaller set of search results after it produced some high-profile errors, but added that the company was still committed to improving the system.

The A.I.-generated summaries are the latest area of tension between tech companies and publishers. The use of articles from news sites has also set off a legal fight over whether companies like OpenAI and Google violated copyright law by taking the content without permission to build their A.I. models.

The New York Times sued OpenAI and its partner, Microsoft, in December, claiming copyright infringement of news content related to the training and servicing of A.I. systems. Seven newspapers owned by Media News Group and Tribune Publishing, including The Chicago Tribune, brought a similar suit against the same tech companies. OpenAI and Microsoft have denied any wrongdoing.

AI Overviews is Google’s latest attempt to catch up to rivals Microsoft and OpenAI, the maker of ChatGPT, in the A.I. race.

More than a year ago, Microsoft put generative A.I. at the heart of its search engine, Bing. Google, afraid to mess with its cash cow, initially took a more cautious approach. But the company announced an aggressive rollout for the A.I. feature at its annual developer conference in mid-May: By the end of the year, more than a billion people would have access to the technology.

AI Overviews combine statements generated from A.I. models with snippets of content from live links across the web. The summaries often contain excerpts from multiple websites while citing sources, giving comprehensive answers without the user ever having to click to another page.

Since its debut, the tool has not always been able to differentiate between accurate articles and satirical posts. When it recommended that users put glue on pizza or eat rocks for a balanced diet, it caused a furor online.

Publishers said in interviews that it was too early to see a difference in traffic from Google since AI Overviews arrived. But the News/Media Alliance, a trade group of 2,000 newspapers, has sent a letter to the Justice Department and the Federal Trade Commission urging the agencies to investigate Google’s “misappropriation” of news content and stop the company from rolling out AI Overviews.

Many publishers said the rollout underscored the need to develop direct relationships with readers, including getting more people to sign up for digital subscriptions and visit their sites and apps directly, and be less reliant on search engines.

Nicholas Thompson, the chief executive of The Atlantic, said his magazine was investing more in all the areas where it had a direct relationship to readers, such as email newsletters.

Newspapers such as The Washington Post and The Texas Tribune have turned to a marketing start-up, Subtext, that helps companies connect with subscribers and audiences through text messaging.

Mike Donoghue, Subtext’s chief executive, said media companies were no longer chasing the largest audiences, but were trying to keep their biggest fans engaged. The New York Post, one of his customers, lets readers exchange text messages with sports reporters on staff as an exclusive subscriber benefit.

Then there’s the dispute over copyright. It took an unexpected turn when OpenAI, which scraped news sites to build ChatGPT, started cutting deals with publishers. It said it would pay companies, including The Associated Press, The Atlantic and News Corp., which owns The Wall Street Journal, to access their content. But Google, whose ad technology helps publishers make money, has not yet signed similar deals. The internet giant has long resisted calls to compensate media companies for their content, arguing that such payments would undermine the nature of the open web.

“You can’t opt out of the future, and this is the future,” said Roger Lynch, the chief executive of Condé Nast, whose magazines include The New Yorker and Vogue. “I’m not disputing whether it will happen or whether it should happen, only that it should happen on terms that will protect creators.”

He said search remained “the lifeblood and majority of traffic” for publishers and suggested that the solution to their woes could come from Congress. He has asked lawmakers in Washington to clarify that the use of content for training A.I. is not “fair use” under existing copyright law and requires a licensing fee.

Mr. Thompson of The Atlantic, whose publication announced a deal with OpenAI on Wednesday, still wishes Google would pay publishers as well. While waiting, he said before the rollout of AI Overviews that despite industry concerns, The Atlantic wanted to be part of Google’s summaries “as much as possible.”

“We know traffic will go down as Google makes this transition,” he said, “but I think that being part of the new product will help us minimize how much it goes down.”

David McCabe contributed reporting.

When Sundar Pichai, Google’s chief executive, introduced a generative artificial intelligence feature for the company’s search engine last month, he and his colleagues demonstrated the new capability with six text-based queries that the public could try out.

The questions included “how do you clean a fabric sofa” and “what should I use to get a coffee stain out of my carpet.” These were intended to highlight how Google’s new feature, A.I. Overviews, could generate full and useful information summaries above traditional search results.

But by Friday, only one of the six queries still yielded an A.I. Overview, according to tests by The New York Times. Instead, the feature was noticeably less prevalent. The search for “what should I use to get a coffee stain out of my carpet” now resulted in a snippet of text from a website, JDog Carpet Cleaning & Floor Care, while “how do you clean a fabric sofa” was replaced by a link to HGTV’s website with the answer. (The results of the searches may vary depending on the user and location.)

The disappearance of A.I. Overviews for some of the searches appeared to be part of a broader rollback after the new technology produced a litany of untruths and errors — including recommending glue as part of a pizza recipe and suggesting that people ingest rocks for nutrients. Users loudly complained on social media about the mistakes, in many cases outright making fun of Google.

Liz Reid, who was recently promoted to Google’s head of search, wrote in a blog post on Thursday that the company had pared back A.I. Overviews in certain ways, launching “additional triggering refinements” to offer more careful responses about health, disabling misleading advice and limiting the inclusion of satire and user responses from forums like Reddit.

“We’ll keep improving when and how we show AI Overviews and strengthening our protections,” she wrote, adding that Google was working on updates to improve broad sets of search results.

Ashley Thompson, a Google spokeswoman, said in a statement on Friday that the company had made more than a dozen technical updates to its systems.

“A.I. Overviews are helping people on a large number of queries on Search today, serving as a jumping-off point to content across the web.” The company added that while it was making adjustments to improve A.I. Overviews, it was not pulling back from the feature for the long term.

The backtracking was a blow to Google’s efforts to keep up with its rivals Microsoft and OpenAI, the maker of the ChatGPT chatbot, in the frenzied race to lead A.I. It also underscored the difficult strategic choice that Google faces over whether to embrace A.I. technology that may not be dependable, or keep its highly popular search engine the same — and risk falling behind its peers.

Google had chosen to go slower than Microsoft, which placed more conversational A.I. into its Bing search engine early last year. Google, which has significantly more users than Bing, tested A.I. features for its search engine a year before introducing A.I. Overviews. The company said the new feature would be rolled out to users in the United States immediately and to more than a billion people by the end of the year.

But ultimately, Google ”should have rolled this out more slowly,” said Patrick Hall, an assistant professor of decision sciences at the George Washington University School of Business. “Once something like this happens, you really have to retreat. And if nothing else, there’s reputational harm” to the company.

Google, which has led internet search for more than two decades, has scrambled since OpenAI released ChatGPT in 2022. Some tech industry insiders considered the chatbot’s ability to generate answers to be a serious threat to Google’s search engine, which has been the most popular way to get information online.

Since then, Google has aggressively worked to regain its advantage in A.I., releasing a family of technology named Gemini, including new A.I. models for developers. The company also infused the technology into YouTube, Gmail and Docs, helping users create videos, emails and drafts with less effort.

Last month, Ms. Reid said at Google’s developer conference that the search engine would do more of the googling for users with A.I. Overviews. She highlighted increasingly complex requests that Google could answer with the feature, but those capabilities have not yet been launched for users.

Onstage, simpler questions like “how do I get the smell of a campfire out of my clothes” yielded A.I. answers including air it out, add baking soda and spray with lemon juice.

But when users got their hands on the new service, they found that A.I. Overviews sometimes generated wrong — or downright dangerous — answers, including recommending putting nontoxic glue on pizza to make the cheese stick. It also misunderstood some websites it was quoting and got presidential history wrong.

Of the six questions that Google launched for the public, the only one that consistently triggered an A.I. Overview on Friday was “what are interesting science projects I can do with my son who is 12 years old.”

The answer? Growing crystals, extracting DNA from saliva and writing a message with invisible ink.

Several major networks cut away from former President Donald J. Trump on Friday during an appearance that had been promoted as a news conference at Trump Tower devolved into a rambling and misleading speech.

It was the latest example of television journalists having to weigh the news value of a major political moment — in this case, the criminal conviction of a former president — against the challenges of reporting on a candidate who regularly speaks in falsehoods.

Mr. Trump’s unfiltered remarks were carried live by cable news channels and NBC, which broke into its usual daytime programming to cover his appearance. In the minutes before he began speaking, MSNBC, CNN and Fox News all aired anticipatory camera shots of an empty lectern.

Mr. Trump began by speaking in his usual discursive, dissembling manner. He unleashed a litany of false statements about his Manhattan trial, attacking witnesses, calling the judge the “devil” and falsely accusing President Biden of being involved in the prosecution.

NBC aired Mr. Trump for 20 minutes before the anchor Lester Holt cut in. “We were told this was going to be a news conference,” he told viewers, before bringing on two legal analysts to dissect and fact-check the remarks. “There is no evidence that Biden was behind any of this,” Mr. Holt said.

ABC and CBS did not interrupt their regular shows.

On MSNBC, where anchors have sometimes refused to air Mr. Trump live, the former president’s appearance aired for about 20 minutes before the network broke away. Later, an on-screen graphic read: “Trump Post-Verdict Remarks Riddled With Lies and Attacks.”

CNN broadcast Mr. Trump for 18 minutes before cutting to a fact-checking segment. Several networks told viewers they would return to Mr. Trump’s appearance once he began speaking with reporters, but the former president did not take press questions. The New York Times, on its website, had a livestream of Mr. Trump’s appearance for about six minutes before cutting the feed and continuing to publish written updates on its blog.

Fox News aired Mr. Trump’s appearance in its entirety.

During the 2016 campaign, Mr. Trump infuriated television journalists when he teased a “major announcement” related to his past lies about Barack Obama’s place of birth. Networks took his remarks live, but the appearance quickly turned into a campaign rally.

“We got played, again, by the Trump campaign,” John King of CNN said at the time.

Elon Musk, Tesla’s chief executive, offered a personal tour of the electric carmaker’s factory in Austin, Texas, to select shareholders this week.

“Please let us know if you have any questions about voting your Tesla shares!” Mr. Musk wrote on X, the social media platform he owns.

It was just one of at least a dozen posts that Mr. Musk has published on X in recent weeks as Tesla’s shareholders have been voting on a $46.5 billion pay package for him.

To encourage approval of the package, Mr. Musk has shared on X a sizzle reel of Tesla’s vehicles speeding through deserts at dusk. He has said he needs enough shares in the company to maintain control over it, especially as it ramps up its artificial intelligence efforts. And he has lashed out at investors who have said they will oppose his pay.

“Thanks to all Tesla vote supporters!” Mr. Musk wrote in a post on May 16, following up two days later with: “Shareholders have the right to vote their shares!” On Thursday, he said shareholders who voted against him were “oathbreakers.”

The messages on X underline how crucial the pay package is for Mr. Musk after a Delaware judge voided it in January. The judge ruled in favor of a dissident shareholder who had sued Tesla, claiming Mr. Musk’s compensation was excessive.

Now Tesla is campaigning to get shareholders to reapprove the pay for Mr. Musk, who has helped build the company into the most valuable automaker in the world. Tesla has been posting on his behalf, too, and the company’s board has publicly supported Mr. Musk’s campaign, saying his performance merits the compensation.

Mr. Musk has turned to his platform of choice, X, to make his case. It’s part of his pattern of increasingly using X to benefit his other companies. In some cases, he has posted support for right-wing heads of state, who have later helped secure advantages for Tesla, including lower tariffs and access to important materials. He also uses the site to market milestones at SpaceX, his rocket company, and the introduction of new vehicles at Tesla to his 185 million followers.

Mr. Musk’s use of X is “a benefit and a curse at the same time,” said Eric Talley, a professor at Columbia Law School. “X is a good way to rally the troops.” But, he added, “you want to have a lawyer making sure he isn’t screwing up his own case.”

Mr. Musk’s posts on X about his Tesla pay package most likely don’t run afoul of the law as long as he isn’t misleading shareholders, corporate governance experts said. But threats like one he posted in January about pursuing robotics and artificial intelligence ventures outside Tesla unless he got 25 percent of the company’s voting shares could be problematic, they added.

In response to a request for comment, a representative for Tesla’s board referred to a post in which Mr. Musk said he didn’t need the money but wanted enough control to ensure that artificial intelligence was handled responsibly. Mr. Musk did not respond to a request for comment, and X declined to comment.

Tesla’s board chair, Robyn Denholm, has posted to a company-backed website advocating for his pay package. “Elon delivered the type of growth that most thought was impossible, and he has created tremendous value for you, the owners of the company,” she wrote.

Tesla’s shareholders first voted on Mr. Musk’s pay package in 2018, approving a plan to grant him an additional 12 percent stake in the company over a dozen years and making him the highest-paid executive in the country. Tesla was valued at $560.2 billion as of the market close on Thursday, and Mr. Musk controls 20.5 percent of it, according to Securities and Exchange Commission filings. (That figure includes shares that have been voided by the Delaware court, and that Tesla is seeking to restore. Without those, his stake is about 13 percent.)

Mr. Musk draws no salary from Tesla. To earn the payouts in company stock, he had to complete ambitious growth milestones at the company.

But Kathaleen McCormick, a judge at the Delaware Chancery Court overseeing the dissident shareholder lawsuit, nullified the pay package, ruling that Mr. Musk held near-total sway over Tesla’s board and essentially approved his own compensation without proper fiduciary management. The judge also ordered him to return his excess pay to Tesla.

In April, Tesla asked shareholders to reapprove Mr. Musk’s pay package. The result will be announced at the company’s annual meeting on June 13.

Mr. Musk often posts about Tesla on X, in part because the carmaker eschews more traditional marketing. He typically hosts splashy online events to debut vehicles or the company’s humanoid robots.

Some of his Tesla posts on X have landed him in trouble. In 2018, the S.E.C. fined Mr. Musk $20 million for claiming on the platform, then known as Twitter, that he planned to take Tesla private at $420 per share. (Tesla paid a separate $20 million fine.) That price, for which he said he had “funding secured,” was 20 percent higher than where Tesla’s stocks were trading at the time. Regulators later said he had misled investors.

As part of his settlement with the S.E.C. in 2018 for the post, Mr. Musk was required to run his social media posts by a company lawyer if the statements contained material information about Tesla. He also stepped down as chairman of Tesla’s board.

Mr. Musk later tried to get out of the settlement, saying it infringed on his freedom of speech. But in 2022, a federal court denied the request. Mr. Musk appealed to the Supreme Court, which declined in April to hear the case.

The S.E.C. declined to comment on Mr. Musk’s public campaign for his pay.

It’s unclear whether the pay package will pass. Some institutional investment firms, like Nordea Asset Management, have come out against the pay package in recent weeks. Tesla shares have fallen about 28 percent this year, and the company is behind schedule on releasing new models. Tesla has also been losing customers to electric carmakers in China.

“Even as Tesla’s performance is floundering, the board has yet to ensure that Tesla has a full-time C.E.O. who is adequately focused on the long-term sustainable success of our company,” a group of institutional investors wrote to shareholders this month. The investor representatives included New York City’s comptroller, Brad Lander, who oversees the city’s pension fund.

Glass Lewis, a proxy advisory firm that consults with institutional investors about how to vote their shares, recommended this week that Tesla shareholders reject Mr. Musk’s package. The firm said that his already sizable ownership in Tesla gave him an incentivize to perform well, and that granting him more shares would dilute the stake of other shareholders.

Glass Lewis opinions are influential with large asset managers, which in Tesla’s case include Vanguard and BlackRock. CalPERS, the California pension fund, also said it would vote against the compensation package.

“Shame on them, they have no honor,” Mr. Musk posted in response on Wednesday.

The chances the pay measure will pass suffered another blow Friday when Institutional Shareholder Services, which also advises institutional investors, recommended against approval.

Even if Tesla’s shareholders vote to reinstate Mr. Musk’s pay, they are unlikely to get final say, legal experts said. The Delaware judge will still need to decide whether the vote is sufficient to reinstate his pay, and the ruling is likely to be appealed.

To survive legal challenges, the pay package needs approval by investors representing more than 50 percent of voting shares not belonging to Mr. Musk or his brother, Kimbal Musk.

Paul Regan, an associate professor at Delaware Law School, said of Tesla’s board: “This thing may end up not going the way they think.”

When OpenAI unveiled the latest version of its immensely popular ChatGPT chatbot this month, it had a new voice possessing humanlike inflections and emotions. The online demonstration also featured the bot tutoring a child on solving a geometry problem.

To my chagrin, the demo turned out to be essentially a bait and switch. The new ChatGPT was released without most of its new features, including the improved voice (which the company told me it postponed to make fixes). The ability to use a phone’s video camera to get real-time analysis of something like a math problem isn’t available yet, either.

Amid the delay, the company also deactivated the ChatGPT voice that some said sounded like the actress Scarlett Johansson, after she threatened legal action, replacing it with a different female voice.

For now, what has actually been rolled out in the new ChatGPT is the ability to upload photos for the bot to analyze. Users can generally expect quicker, more lucid responses. The bot can also do real-time language translations, but ChatGPT will respond in its older, machine-like voice.

Nonetheless, this is the leading chatbot that upended the tech industry, so it was worth reviewing. After trying the sped-up chatbot for two weeks, I had mixed feelings. It excelled at language translations, but it struggled with math and physics. All told, I didn’t see a meaningful improvement from the last version, ChatGPT-4. I definitely wouldn’t let it tutor my child.

This tactic, in which A.I. companies promise wild new features and deliver a half-baked product, is becoming a trend that is bound to confuse and frustrate people. The $700 Ai Pin, a talking lapel pin from the start-up Humane, which is funded by OpenAI’s chief executive, Sam Altman, was universally panned because it overheated and spat out nonsense. Meta also recently added to its apps an A.I. chatbot that did a poor job at most of its advertised tasks, like web searches for plane tickets.

Companies are releasing A.I. products in a premature state partly because they want people to use the technology to help them learn how to improve it. In the past, when companies unveiled new tech products like phones, what we were shown — features like new cameras and brighter screens — was what we were getting. With artificial intelligence, companies are giving a preview of a potential future, demonstrating technologies that are being developed and working only in limited, controlled conditions. A mature, reliable product might arrive — or might not.

The lesson to learn from all this is that we, as consumers, should resist the hype and take a slow, cautious approach to A.I. We shouldn’t be spending much cash on any underbaked tech until we see proof that the tools work as advertised.

The new version of ChatGPT, called GPT-4o (“o” as in “omni”), is now free to try on OpenAI’s website and app. Nonpaying users can make a few requests before hitting a timeout, and those who have a $20 monthly subscription can ask the bot a larger number of questions.

OpenAI said its iterative approach to updating ChatGPT allowed it to gather feedback to make improvements.

“We believe it’s important to preview our advanced models to give people a glimpse of their capabilities and to help us understand their real-world applications,” the company said in a statement.

(The New York Times sued OpenAI and its partner, Microsoft, last year for using copyrighted news articles without permission to train chatbots.)

Here’s what to know about the latest version of ChatGPT.

Geometry and Physics

To show off ChatGPT-4o’s new tricks, OpenAI published a video featuring Sal Khan, the chief executive of the Khan Academy, the education nonprofit, and his son, Imran. With a video camera pointed at a geometry problem, ChatGPT was able to talk Imran through solving it step by step.

Even though ChatGPT’s video-analysis feature has yet to be released, I was able to upload photos of geometry problems. ChatGPT solved some of the easier ones correctly, but it tripped up on more challenging problems.

For one problem involving intersecting triangles, which I dug up on an SAT preparation website, the bot understood the question but gave the wrong answer.

Taylor Nguyen, a high school physics teacher in Orange County, Calif., uploaded a physics problem involving a man on a swing that is commonly included on Advanced Placement Calculus tests. ChatGPT made several logical mistakes to give the wrong answer, but it was able to correct itself with feedback from Mr. Nguyen.

“I was able to coach it, but I’m a teacher,” he said. “How is a student supposed to pick out those mistakes? They’re making this assumption that the chatbot is right.”

I did notice that ChatGPT-4o succeeded at some division calculations that its predecessors did incorrectly, so there are signs of slow improvement. But it also failed at a basic math task that past versions and other chatbots, including Meta AI and Google’s Gemini, have flunked at: the ability to count. When I asked ChatGPT-4o for a four-syllable word starting with the letter “W,” it responded, “Wonderful.”

OpenAI said it was constantly working to improve its systems’ responses to complex math problems.

Mr. Khan, whose company uses OpenAI’s technology in its tutoring software Khanmigo, did not respond to a request for comment on whether he would leave ChatGPT the tutor alone with his son.

Reasoning

OpenAI also highlighted that the new ChatGPT was better at reasoning, or using logic to come up with responses. So I ran it through one of my favorite tests: I asked it to generate a Where’s Waldo? puzzle. When it showed an image of a giant Waldo standing in a crowd, I said that the point is that he’s supposed to be hard to find.

The bot then generated an even larger Waldo.

Subbarao Kambhampati, a professor and researcher of artificial intelligence at Arizona State University, also put the chatbot through some tests and said he saw no noticeable improvement in reasoning compared with the last version.

He presented ChatGPT a puzzle involving blocks:

If block C is on top of block A, and block B is separately on the table, can you tell me how I can make a stack of blocks with block A on top of block B and block B on top of block C, but without moving block C?

The answer is that it’s impossible to arrange the blocks under these conditions, but, just as with past versions, ChatGPT-4o consistently came up with a solution that involved moving block C. With this and other reasoning tests, ChatGPT was occasionally able to take feedback to get the correct answer, which is antithetical to how artificial intelligence is supposed to work, Mr. Kambhampati said.

“You can correct it, but when you do that you’re using your own intelligence,” he said.

OpenAI pointed to test results that showed GPT-4o scored about two percentage points higher at answering general knowledge questions than previous versions of ChatGPT, illustrating that its reasoning skills had slightly improved.

Language

OpenAI also said the new ChatGPT could do real-time language translation, which could help you converse with someone speaking a foreign language.

I tested ChatGPT with Mandarin and Cantonese and confirmed that it was OK at translating phrases, such as “I’d like to book a hotel room for next Thursday” and “I want a king-size bed.” But the accents were slightly off. (To be fair, my broken Chinese is not much better.) OpenAI said it was still working to improve accents.

ChatGPT-4o also excelled as an editor. When I fed it paragraphs that I wrote, it was fast and effective at removing excessive words and jargon. ChatGPT’s decent performance with language translation gives me confidence that this will soon become a more useful feature.

Bottom Line

A major thing OpenAI got right with ChatGPT-4o is making the technology free for people to try. Free is the right price: Since we are helping to train these A.I. systems with our data to improve, we shouldn’t be paying for them.

The best of A.I. has yet to come, and it might one day be a good math tutor that we want to talk to. But we should believe it when we see it — and hear it.

The website calling itself Man Stuff News caters to a certain sensibility, with categories like “Backyard Grilling,” “TV Shows for Guys” and “Beard Grooming.” A recent article headlined “Tips for Dads During Labor” offered this nugget of advice: “Just remember to spend some time together before deciding whether or not to give birth.”

Get to its section devoted to world news, however, and the nature of the coverage changes drastically. There, a recent article belittled an international warrant to arrest Russia’s president, Vladimir V. Putin, for war crimes. It repeated, word for word, an article that had appeared a day before under a different byline on the website for RT, Russia’s global television network.

RT, which the U.S. State Department describes as a key player in the Kremlin’s disinformation and propaganda apparatus, has been blocked in the European Union, Canada and other countries since Russia invaded Ukraine in 2022. Sites like Man Stuff News, however, have helped RT sidestep the restrictions and continue reaching European and American audiences, according to a new report.

Replicas of RT articles have been laundered thousands of times through hundreds of sites, according to the report, written by researchers from the German Marshall Fund, the University of Amsterdam and the Institute for Strategic Dialogue, a research nonprofit. The sites include content aggregators like Infowars, run by the conspiracy theorist Alex Jones; mirrors of RT repurposed from abandoned “zombie” sites; faux local news outlets with names like San Francisco Telegraph; and domains focusing on spirituality, yoga, extraterrestrials and the apocalypse. Many of the articles were then further disseminated through social media.

The rationale for reposting RT content most likely varies from site to site, but the surreptitious republishing represents a particular danger in the European Union, where concerns about Kremlin-linked disinformation campaigns are intensifying, especially as Russia tries to weaken European support for Ukraine ahead of parliamentary elections next week.

“This is really the tip of the Russian propaganda iceberg,” said Bret Schafer, a co-author of the report and a senior fellow at German Marshall Fund. “It was quite evident when we were running the search results in the E.U. that if Russian propaganda is not showing up on Russian domains, it’s getting through, which is sort of a double whammy because it’s not just evading restrictions and bans, it’s doing so on sites that are less transparent than RT itself.”

RT said in a statement that its content did not follow the “U.S. State Department/NATO party line” and added that it is “very glad that RT’s news content is so massively popular with a wide range of platforms and users.”

A message sent to an email address listed for the Man Stuff News website registration went unanswered. The site offers few details about where it is based or who operates it.

As non-Russian sources parroted the Kremlin’s talking points, they helped legitimize the narratives to an often unsuspecting audience, the researchers concluded. The copied articles, which the researchers described as “Russia’s propaganda nesting dolls,” targeted a huge geographic swath of viewers via sites registered in at least 40 countries across six continents, including in countries where RT is ostensibly blocked. When factoring in RT’s content in languages other than English along with other Kremlin-controlled media outlets, the true scope of Russian propaganda laundering is probably much greater, the researchers said.

Ursula von der Leyen, the president of the European Commission, said in a speech this month that she was “particularly concerned about the rise of foreign interference and manipulation in our societies, our democracies and our elections.” She cited “swarms of negative disinformation” about specific issues and candidates and malicious attempts at “buying influence and causing chaos.”

Last month, a consortium of 36 European fact-checking organizations said that false or misleading content about the European Union or Ukraine was among the most prevalent forms of disinformation it had encountered.

An E.U. report this year said operatives abroad — most obviously from Russia, but also China — were coordinating on “virtually all platforms” to create an alternative information environment that would erode trust in democracy. Last month, the European Commission conducted a pre-election stress test to evaluate the platforms’ preparedness against A.I.-generated fakery, influence campaigns from bot accounts and other threats.

Since 2022, the Kremlin has been unable to gain access to some of its main messaging channels in the West after Canada and the European Union took RT off their airwaves. This month, the bloc suspended four other Russian media outlets from broadcasting.

In the United States, government regulators did not take action against the Russian network’s American outpost, RT America. Instead, television distributors across the country cut ties with RT America in early 2022, and it shut down within days.

Online platforms have also tried to curb RT’s reach; YouTube blocked global access to channels affiliated with RT and said it strove to remove harmful misinformation. Laundered RT content, however, persists there and on other platforms, researchers said, echoing findings from other research groups. On YouTube, RT articles seemed to have been narrated using an automated text-to-speech generator to evade filters. Content copied from RT also appeared on major social and messaging sites like Facebook, LinkedIn, Pinterest, Substack, Telegram and X, as well as on niche platforms such as Gab and Rumble, the researchers said.

Working from more than 1,500 RT articles published last year, the researchers looked for websites that featured similar content or metadata, limiting their search to results geolocated in the United States and Belgium, the European Union’s de facto capital.

Some of the sites were probably circulating RT’s content with the network’s permission, the researchers said, while others had plagiarized RT without its knowledge. The sites may have been ideologically aligned with the Kremlin, or more intent on driving traffic to boost visibility or ad revenue. Some of the sites disclosed that they were reposting RT content. (Man Stuff News ended its copy of the article about Mr. Putin’s arrest warrant by posting the web address of the original RT story.)

Verbatim replicas of RT articles appeared on media outlets affiliated with governments in Cambodia, Iran, Nigeria and Yemen, as well as on one Lebanese outlet owned by Hezbollah, an Iranian-backed Lebanese militia. Researchers connected one website to a conservative Catholic online ministry in Texas that had posts about abortion, candlemaking and, in an example lifted from RT, the lack of aid after an earthquake in Syria.

Researchers noted that RT was far from the only Kremlin media outlet being laundered. As major elections approach in the European Union and the United States, Russian disinformation operatives have honed their strategies. Recent videos featuring synthetic voices and other signs of manipulation by artificial intelligence targeted right-wing American voters with fake messages about President Biden. Fake news organizations crafted by Russian operatives have mimicked actual American outlets while promoting Kremlin propaganda; one former sheriff’s deputy in Florida who received political asylum in Moscow has built more than 160 such fake sites.

X, the social media service owned by Elon Musk, will host live video town halls with former President Donald J. Trump and Robert F. Kennedy Jr. in the run-up to the U.S. presidential election, a major push into politics, according to a person familiar with the plans.

The town halls, which will be produced in a partnership with the cable channel NewsNation, have yet to be scheduled and moderators have not been selected, said the person, who declined to be named because the plans were not yet public. X users will be able to ask questions of the presidential candidates during the events.

In an appearance on NewsNation on Wednesday night, Mr. Kennedy confirmed his plans to participate in a town hall, adding that Mr. Musk offered the use of the platform.

X has undergone a transformation under Mr. Musk’s ownership, as the billionaire has drastically reduced its work force, overhauled its technology and used the platform to advance his personal agendas.

The town halls follow a March meeting between Mr. Trump and Mr. Musk. The X owner has posted about President Biden more frequently in recent months, criticizing him on a range of fronts, including his age and his policies on immigration. X invited Mr. Biden to hold a town hall on the platform, but his campaign has not responded, the person familiar with the arrangement said.

Mr. Musk has stopped short of endorsing Mr. Trump, who was barred from Twitter after the Capitol riot on Jan. 6, 2021. Mr. Musk quickly reinstated the former president after buying the platform in 2022, which he renamed X. Mr. Musk has courted Mr. Trump, asking him to return to using X.

Mr. Trump’s campaign did not immediately respond to a request for comment.

Mr. Trump’s presence on X — if only for the town hall — could bring fresh viewership to the platform, which has struggled under Mr. Musk’s ownership. Advertisers have retreated, concerned about reports of hate speech and misinformation on the platform.

The platform has long been known as a text-first social media service. But under its chief executive, Linda Yaccarino, a former NBC Universal advertising leader, the company has increasingly incorporated video features.

She has made several deals to produce video content with TV personalities including Tucker Carlson, the former Fox News host; Tulsi Gabbard, the former Democratic presidential candidate; and Paris Hilton, the reality star and D.J.

Ms. Yaccarino approached Nexstar, the parent company of NewsNation, about the town hall concept, the person familiar with the arrangement said, and has been ironing out the details as part of her push to make X a video platform. Axios previously reported the town hall.

At a sprawling complex in Warren, Mich., General Motors’ hopes for its driverless car future play out in a virtual reality headset offered to visitors.

In a video, the electric and autonomous car drives itself. Wirelessly connected to traffic lights and the surrounding streets, the car avoids collisions and reduces congestion, part of what G.M. calls its “0-0-0” vision — “zero crashes, zero emission, zero congestion.”

At least, that’s the plan. G.M.’s driverless future looks a lot further away today than it did a year ago, when Cruise, G.M.’s driverless car subsidiary, was deep into an aggressive expansion of its robot taxi services, testing in 15 cities across 10 states.

On Oct. 2, a Cruise driverless car hit and dragged a pedestrian for 20 feet on a San Francisco street, causing severe injuries. Weeks later, the California Department of Motor Vehicles accused Cruise of omitting the dragging from a video of the incident that was initially provided to the agency and suspended the company’s license in the state.

In November, Cruise voluntarily paused all operations across the country after facing widespread criticism that it was neglecting safety as it expanded its driverless taxi service. Cruise also pushed out nine executives, its chief executive stepped down, and the company laid off a quarter of its work force.

Now comes the hard part: Rebuilding a ruined reputation. In recent interviews with The New York Times, the three executives now running Cruise say they are in no rush to get back on the road. After learning the hard way about the risks of moving too fast with a cutting-edge technology, Cruise has slowed its breakneck development to a crawl to avoid another major mishap.

“For a long time before, Cruise was really moving fast and other competitors were not,” said Craig Glidden, who became president and chief administrative officer of Cruise in November. Now, he said, safety is Cruise’s “North Star.”

But going slow means the company risks falling far behind its top rivals. Waymo, a subsidiary of Google’s parent company, Alphabet, has had driverless taxis operating in the Phoenix area since 2020 and San Francisco since late 2022 without serious incidents, and it recently expanded to Los Angeles. Zoox, an Amazon subsidiary, has been testing a steering-wheel-free robot taxi in Las Vegas since last June.

“Catching up with Waymo technologically is going to take three to five years at best,” said Alex Roy, a consultant and former executive in the autonomous car industry. He added that it was even harder for Cruise to catch up commercially because Waymo was “generating revenues with trust that Cruise never earned.”

Some industry observers were surprised G.M. didn’t shut down Cruise after its public meltdown late last year. Since acquiring the company in 2016, G.M. has spent over $8 billion on its driverless subsidiary. Cruise lost $3.48 billion last year, and another $519 million over the first three months of 2024.

“I was thinking in the late part of 2023 and into early 2024 that the most likely outcome was that they were going to completely turn off Cruise,” said Reilly Brennan, a partner at Trucks Venture Capital, which invests in the future of transportation.

But after slashing $1 billion from Cruise’s 2024 budget, Mary T. Barra, G.M.’s chief executive, reiterated her commitment to the company during earnings calls. In April, she told investors that Cruise had made “tangible progress,” although G.M. is exploring different options to fund the business, including taking outside investments.

After Cruise’s former chief executive and co-founder Kyle Vogt resigned in November, G.M. appointed two presidents who report to its board: Mo Elshenawy, previously the company’s executive vice president of engineering, and Mr. Glidden, who also serves as G.M.’s general counsel. In February, Cruise hired Steve Kenner, a veteran product safety executive, as chief safety officer.

The three executives all decide on safety decisions, such as when to take the next step in deployment. Those calls, Mr. Kenner said, have to be unanimous.

So far, Cruise has taken baby steps back to the road. In April, it picked Phoenix, the home to its operations center, to be the first city to restart testing with human drivers. On May 13, after a month of driving a handful of vehicles in order to understand local road features, Cruise transitioned into supervised autonomous testing, with two safety drivers per vehicle.

Cruise used to say its robot taxis were, on average, safer than a human driver. But so-called edge cases — incidents like road construction or erratic cyclists that humans can intuitively react to — bedeviled the robot taxis. Mr. Elshenawy said the cars had improved their navigation of construction zones and how they deal with emergency vehicles.

Cruise hopes to offer driverless ride-hailing service in one city by the end of 2024, while operating with safety drivers in fewer than five cities, Mr. Glidden said. That is, if the edge case issue can be improved.

While Mr. Elshenawy’s engineering team works to improve the technology, Mr. Glidden and Mr. Kenner have been traveling across the country to meet with regulators. Cruise has met with local officials and state regulators in Arizona, Texas and California, as well as with the National Highway Traffic Safety Administration. It has also spoken with several cities in the Southeast where it previously tested its fleet.

In California, Cruise has answered questions from state regulators about driverless testing, but it is unclear if or when it could regain a permit. The talent pool in Silicon Valley is essential to Cruise’s business, so executives say they are committed to staying in the state.

Whether Cruise’s cautious approach restores faith in the company among regulators is an open question. Dave Cortese, a California state senator representing Silicon Valley, said the autonomous vehicle industry’s aggressive testing on public roads in the past had “created tension and distrust.”

For the company to win over regulators, it needs a “profound demonstration of transparency” to demonstrate that an incident like Oct. 2 will not happen again, said Mr. Roy, the consultant.

“We may not agree, but I think there are lots of places where we do agree,” said Tilly Chang, executive director of San Francisco County Transportation Authority. “But it is also unclear to us what it would take for them to get reinstated.”

When Mark Zuckerberg, the chief executive of Meta, announced last year that his company would release an artificial intelligence system, Jeffrey Emanuel had reservations.

Mr. Emanuel, a part-time hacker and full-time A.I. enthusiast, had tinkered with “closed” A.I. models, including OpenAI’s, meaning the systems’ underlying code could not be accessed or modified. When Mr. Zuckerberg introduced Meta’s A.I. system by invitation only to a handful of academics, Mr. Emanuel was concerned that the technology would remain limited to just a small circle of people.

But in a release last summer of an updated A.I. system, Mr. Zuckerberg made the code “open source” so that it could be freely copied, modified and reused by anyone.

Mr. Emanuel, the founder of the blockchain start-up Pastel Network, was sold. He said he appreciated that Meta’s A.I. system was powerful and easy to use. Most of all, he loved how Mr. Zuckerberg was espousing the hacker code of making the technology freely available — largely the opposite of what Google, OpenAI and Microsoft have done.

“We have this champion in Zuckerberg,” Mr. Emanuel, 42, said. “Thank God we have someone to protect the open-source ethos from these other big companies.”

Mr. Zuckerberg has become the highest-profile technology executive to support and promote the open-source model for A.I. That has put the 40-year-old billionaire squarely on one end of a divisive debate over whether the potentially world-changing technology is too dangerous to be made available to any coder who wants it.

Microsoft, OpenAI and Google have more of a closed A.I. strategy to guard their tech, out of what they say is an abundance of caution. But Mr. Zuckerberg has loudly stood behind how the technology should be open to all.

“This technology is so important, and the opportunities are so great, that we should open source and make it as widely available as we responsibly can, so that way everyone can benefit,” he said in an Instagram video in January.

That stance has turned Mr. Zuckerberg into the unlikely man of the hour in many Silicon Valley developer communities, prompting talk of a “glow-up” and a kind of “Zuckaissance.” Even as the chief executive continues grappling with scrutiny over misinformation and child safety issues on Meta’s platforms, many engineers, coders, technologists and others have embraced his position on making A.I. available to the masses.

Since Meta’s first fully open-source A.I. model, called LLaMA 2, was released in July, the software has been downloaded more than 180 million times, the company said. A more powerful version of the model, LLaMA 3, which was released in April, reached the top of the download charts on Hugging Face, a community site for A.I. code, at record speed.

Developers have created tens of thousands of their own customized A.I. programs on top of Meta’s A.I. software to perform everything from helping clinicians read radiology scans to creating scores of digital chatbot assistants.

“I told Mark, I think that open sourcing LLaMA is the most popular thing that Facebook has done in the tech community — ever,” said Patrick Collison, chief executive of the payments company Stripe, who recently joined a Meta strategic advisory group that is aimed at helping the company make strategic decisions about its A.I. technology. Meta owns Facebook, Instagram and other apps.

Mr. Zuckerberg’s new popularity in tech circles is striking because of his fraught history with developers. Over two decades, Meta has sometimes pulled the rug out from under coders. In 2013, for instance, Mr. Zuckerberg bought Parse, a company that built developer tools, to attract coders to build apps for Facebook’s platform. Three years later, he shuttered the effort, angering developers who had invested their time and energy in the project.

A spokeswoman for Mr. Zuckerberg and Meta declined to comment. (The New York Times last year sued OpenAI and its partner, Microsoft, claiming copyright infringement of news content related to A.I. systems.)

Open-source software has a long and storied history in Silicon Valley, with major tech battles revolving around open versus proprietary — or closed — systems.

In the internet’s early days, Microsoft jockeyed to provide the software that ran internet infrastructure, only to eventually lose out to open-source software projects. More recently, Google open sourced its Android mobile operating system to take on Apple’s closed iPhone operating system. Firefox, the internet browser, WordPress, a blogging platform, and Blender, a popular set of animation software tools, were all built using open-source technologies.

Mr. Zuckerberg, who founded Facebook in 2004, has long backed open-source technology. In 2011, Facebook started the Open Compute Project, a nonprofit that freely shares designs of servers and equipment inside data centers. In 2016, Facebook also developed Pytorch, an open-source software library that has been widely used to create A.I. applications. The company is also sharing blueprints of computing chips that it has developed.

“Mark is a great student of history,” said Daniel Ek, Spotify’s chief executive, who considers Mr. Zuckerberg a confidant. “Over time in the computing industry, he’s seen that there’s always been closed and open paths to take. And he has always defaulted to open.”

At Meta, the decision to open source its A.I. was contentious. In 2022 and 2023, the company’s policy and legal teams supported a more conservative approach to releasing the software, fearing a backlash among regulators in Washington and the European Union. But Meta technologists like Yann LeCun and Joelle Pineau, who spearhead A.I. research, pushed the open model, which they argued would better benefit the company in the long term.

The engineers won. Mr. Zuckerberg agreed that if the code was open, it could be improved and safeguarded faster, he said in a post last year on his Facebook page.

While open sourcing LLaMA means giving away computer code that Meta spent billions of dollars to create with no immediate return on investment, Mr. Zuckerberg calls it “good business.” As more developers use Meta’s software and hardware tools, the more likely they are to become invested in its technology ecosystem, which helps entrench the company.

The technology has also helped Meta improve its own internal A.I. systems, aiding ad targeting and recommendations of more relevant content on Meta’s apps.

“It is 100 percent aligned with Zuckerberg’s incentives and how it can benefit Meta,” said Nur Ahmed, a researcher at MIT Sloan who studies A.I. “LLaMA is a win-win for everybody.”

Competitors are taking note. In February, Google open sourced the code for two A.I. models, Gemma 2B and Gemma 7B, a sign that it was feeling the heat from Mr. Zuckerberg’s open-source approach. Google did not respond to requests for comment. Other companies, including Microsoft, Mistral, Snowflake and Databricks, have also started offering open-source models this year.

For some coders, Mr. Zuckerberg’s A.I. approach hasn’t erased all of the baggage of the past. Sam McLeod, 35, a software developer in Melbourne, Australia, deleted his Facebook accounts years ago after growing uncomfortable with the company’s track record on user privacy and other factors.

But more recently, he said, he recognized that Mr. Zuckerberg had released “cutting edge” open-source software models with “permissive licensing terms,” something that can’t be said for other big tech companies.

Matt Shumer, 24, a developer in New York, said he had used closed A.I. models from Mistral and OpenAI to power digital assistants for his start-up, HyperWrite. But after Meta released its updated open-source A.I. model last month, Mr. Shumer started relying heavily on that instead. Whatever reservations he had about Mr. Zuckerberg are in the past.

“Developers have started to see past a lot of issues they’ve had with him and Facebook,” Mr. Shumer said. “Right now, what he’s doing is genuinely good for the open-source community.”

A dozen years ago, John Mark Dougan, a former deputy sheriff in Palm Beach County, Fla., sent voters an email posing as a county commissioner, urging them to oppose the re-election of the county’s sheriff.

He later masqueraded online as a Russian tech worker with a pseudonym, BadVolf, to leak confidential information in violation of state law, fooling officials in Florida who thought they were dealing with a foreigner.

He also posed as a fictional New York City heiress he called Jessica, tricking an adviser to the Palm Beach County Sheriff’s Office into divulging improper conduct by the department.

“And boy, did he ever spill ALL of the beans,” Mr. Dougan said in a written response to questions for this article, in which he confirmed his role in these episodes.

Those subterfuges in the United States, it turned out, were only a prelude to a more prominent and potentially more ominous campaign of deception he has been conducting from Russia.

Mr. Dougan, 51, who received political asylum in Moscow, is now a key player in Russia’s disinformation operations against the West. Back in 2016, when the Kremlin interfered in the American presidential election, an army of computer trolls toiled for hours in an office building in St. Petersburg to try to fool Americans online.

Today Mr. Dougan may be accomplishing much the same task largely by himself, according to American and European government officials and researchers from companies and organizations that have tracked his activities since August. The groups include NewsGuard, a company that reviews the reliability of news and information online; Recorded Future, a threat intelligence company; and Clemson University’s Media Forensics Hub.

Working from an apartment crowded with servers and other computer equipment, Mr. Dougan has built an ever-growing network of more than 160 fake websites that mimic news outlets in the United States, Britain and France.

With the help of commercially available artificial intelligence tools, including OpenAI’s ChatGPT and DALL-E 3, he has filled the sites with tens of thousands of articles, many based on actual news events. Interspersed among them are also bespoke fabrications that officials in the United States and European Union have attributed to Russian intelligence agencies or the administration of President Vladimir V. Putin.

Between September and May, Mr. Dougan’s outlets have been cited or referred to in news articles or social media posts nearly 8,000 times, and seen by more than 37 million people in 16 languages, according to a report released Wednesday by NewsGuard.

The fakes have recently included a baseless article on a fake San Francisco Chronicle website that said Ukraine’s president, Volodymyr Zelensky, had smuggled 300 kilograms of cocaine from Argentina. Another false narrative appeared last month in the sham Chronicle and on another site, called The Boston Times, claiming that the C.I.A. was working with Ukrainians to undermine Donald J. Trump’s presidential campaign.

Mr. Dougan, in a series of text exchanges and one telephone interview with The New York Times, denied operating the sites. A digital trail of clues, including web domains and internet protocol addresses, suggests otherwise, the officials and researchers say.

A friend in Florida who has known Mr. Dougan for 20 years, Jose Lambiet, also said in a telephone interview that Mr. Dougan told him in January that he had created the sites.

Steven Brill, a founder of NewsGuard, which has spent months tracking Mr. Dougan’s work, said he represented “a massive incursion into the American news ecosystem.”

“It’s not just some guy sitting in his basement in New Jersey tapping out a phony website,” he added.

Mr. Dougan’s emergence as a weapon of the Kremlin’s propaganda war follows a troubled life in the United States that included home foreclosures and bankruptcy. As a law enforcement officer in Florida and Maine, he faced accusations of excessive use of force and sexual harassment that resulted in costly lawsuits against the departments he worked for.

He faces an arrest warrant in Florida — its records sealed by court order — on 21 felony charges of extortion and wiretapping that resulted from a long-running feud with the sheriff of Palm Beach County.

Mr. Dougan’s activities from Moscow, where he fled in 2016 one step ahead of those charges, continue to draw scrutiny from the authorities in the United States. Last year, he impersonated an F.B.I. agent in a telephone call to Mr. Brill, according to an account by Mr. Brill to be published next week in a new book, “The Death of Truth.”

Mr. Dougan, who acknowledged making the call in a text message this week, had been angered by a NewsGuard report in February 2023 that criticized YouTube for allowing videos parroting Russian propaganda about the war in Ukraine, including some by Mr. Dougan.

In a rambling, profanity-laced video in response on YouTube last year, Mr. Dougan posted excerpts from the call with Mr. Brill and showed a Google Earth satellite photograph of his home in Westchester County, a suburb of New York City — “just down the road from the Clinton crime family,” as Mr. Dougan put it, referring to the home of former President Bill Clinton and former Secretary of State Hillary Rodham Clinton.

The call prompted an F.B.I. investigation that, according to Mr. Brill, traced the call to Mr. Dougan’s telephone in Russia. (A spokeswoman for the bureau did not respond to a request for comment on the investigation or Mr. Dougan’s previous activities.)

A History in Law Enforcement

Mr. Dougan began to hone the skills that he is putting to use today during a turbulent childhood in the United States. In the written responses to questions for this article, he said he had struggled at home and in school, bullied because of Tourette’s syndrome, but found a passion in computers. When he was 8, he said, the man who would become his stepfather began teaching him to write computer code.

“By the time I was 16,” he wrote in one response, “I knew a dozen different programming languages.”

After a four-year stint in the Marine Corps, which he claims he offered to join in lieu of a jail sentence for fleeing a police stop for speeding on a motorcycle, he became a police officer first in a small force in Mangonia Park, Fla., and then the Palm Beach County Sheriff’s Office from 2005 to 2009.

According to news reports and his own accounts over the years, Mr. Dougan repeatedly clashed with superiors and colleagues, facing numerous internal investigations that he said were retaliatory because he objected to police misconduct, including instances of racial bias.

In 2009, he moved briefly to Windham, Maine, to work in another small-town police department. There he faced a complaint of sexual harassment that resulted in his dismissal before he completed his probationary period.

Mr. Dougan started a website called WindhamTalk to defend himself. The website foreshadowed others he would create, including one devoted to the Palm Beach County Sheriff’s Office, PBSOTalk.

After moving back to Florida, he used PBSOTalk to torment in particular the department’s elected sheriff, Ric L. Bradshaw, whom he accused of corruption. He posted the unlawful recordings of “Jessica” chatting with a former detective commander, Mark Lewis, who, Mr. Dougan claimed, was investigating the sheriff’s critics, including himself. As Mr. Dougan acknowledged in a video interview last year, it is illegal in Florida to record a telephone conversation without permission.

In a statement, a spokeswoman for the Sheriff’s Office, Therese C. Barbera, said Mr. Dougan was “a wanted felon for cyberstalking using unsubstantiated and fabricated claims that have NO factual basis.”

In February 2016, PBSOTalk posted confidential information about thousands of police officers, federal agents and judges. The next month, F.B.I. agents and local police officers searched Mr. Dougan’s home, seizing all of his electronic equipment.

Fearing arrest, he said, he made his way to Canada and caught a flight to Moscow. He was indicted on the 21 Florida felony charges the next year.

Peddling Russia Propaganda

In Russia, Mr. Dougan refashioned himself as a kind of journalist, documenting his travels around the country, including Lake Baikal in Siberia and Crimea, the peninsula in Ukraine that Russia annexed in 2014 in violation of international law.

He posted photographs and videos from those trips on YouTube, which suspended his channel after NewsGuard’s report last year. He also appeared regularly on state media, including with two former intelligence operatives, Maria Butina, who penetrated Republican political circles, and Anna Chapman, one of 10 spies who inspired the television series “The Americans.”

In 2021, as Mr. Putin began mobilizing the military forces that would invade Ukraine, Mr. Dougan posted a video that the Kremlin would cite as one justification for its attack. In it, he claimed that the United States operated biological weapons factories in Ukraine, an accusation that Russia and its allies have pushed without ever providing evidence.

Once the war started, Mr. Dougan recounted in his written responses to questions, he traveled to Ukraine 14 times to report from the Russian side of the front lines. He appeared in Russian government hearings purporting to expose Ukraine’s transgressions, indicating some level of cooperation with the government authorities.

He has faced criticism for the reports, including in a profile in The Daily Beast, that he posted on YouTube and other platforms. Mr. Dougan has portrayed the war much as Russia’s propaganda has: as a righteous battle against neo-Nazis backed by a decadent West, led by the United States and NATO.

“The West has consistently lied about every aspect of this conflict,” he wrote. “Why does only one side get to tell their story?”

Fake News Sites in the U.S.

In April 2021, Mr. Dougan revived a website called DC Weekly, which had been created four years earlier and published fake articles about the Palm Beach County Sheriff’s Office. According to a report last December by Clemson’s Media Forensics Hub, the domain and internet protocol address were shared by PBSOTalk and Mr. Dougan’s personal website, as well as two marketing books he wrote in exile and a security firm he operated, Falcon Eye Tech, which offered “offshore security monitoring services.”

After Russia’s assault on Ukraine began in 2022, the site carried articles about the war.

Then, last August, the site began to publish articles based on elaborate fabrications that the Western government officials and disinformation researchers said came from Russia’s propaganda units. They often appeared first in videos or audio recordings on obscure X accounts or YouTube channels, then spread to sites like DC Weekly and then to Russian state media as if they were authentic accusations, a process researchers call “narrative laundering.”

The baseless narratives included claims that relatives or cronies of Ukraine’s leader secretly bought luxury properties, yachts or jewelry, and that Prince Andrew, the brother of King Charles III of Britain, had abducted and abused children during a secret visit to Ukraine.

Dozens of new sites have appeared in recent months. They included ones made to look like local news outlets: The Chicago Chronicle, The Miami Chronicle, The Boston Times, The Flagstaff Post and The Houston Post. Some hijacked names of actual news organizations, like The San Francisco Chronicle, or approximated them, in the case of one called The New York News Daily.

When The New York Times reported on the new sites in March, DC Weekly published a lengthy response in a stilted style that indicated the use of artificial intelligence. It was written under the name Jessica Devlin, one of the fictitious journalists on the site. “I’m not a shadowy foreign actor,” the article said.

At the end, the article invited media inquiries at an email address with the domain Falcon Eye Tech.

Two days later, Mr. Dougan answered.

103 New Sites in Two Days

Mr. Dougan, who became a Russian citizen last year and voted in the country’s presidential election in March, said in his messages to The Times that he made a living by selling security devices he designed for a manufacturer in China. He denied being paid by any Russian authorities, claiming he funds his activities himself.

His friend Mr. Lambiet, a private investigator and former journalist, said he considered Mr. Dougan a good man but cautioned that Mr. Dougan had a propensity to make things up. “He’s like a Russian disinformation campaign: It’s hard to know what’s true and what’s not,” he said.

As evidence of Mr. Dougan’s role in the news sites has emerged, he has shifted tactics. Recorded Future, the threat intelligence company, released a report this month that detailed his ties to agencies linked to the Russian disinformation. The report documented the extensive use of A.I., which one of the company’s researchers, Clément Briens, estimated made Mr. Dougan’s work far cheaper than hiring a troll army.

At the time, Recorded Future identified 57 domains that Mr. Dougan had created. In a two-day span after the report was published, 103 new sites appeared, all on a server in California.

“He’s trying to obfuscate the Russian links,” Mr. Briens said.

Mr. Dougan at times treats his activities as a game of cat and mouse. He spent months engaging with a researcher at NewsGuard, McKenzie Sadeghi, revealing details of his life in Moscow while mocking her boss, Mr. Brill.

“He seemed to be toying with me, both to elicit my responses and, it seemed, to show off his talent for global online mischief, without actually admitting anything,” she wrote in the report published on Wednesday.

While Mr. Dougan’s sites have focused on Russian narratives about the war in Ukraine, the researchers and government officials say he has laid the foundation for interference in the unusually large confluence of elections taking place around the world this year.

This suggests a “risk of an expanded operation scope in the near future, potentially targeting diverse audiences and democratic systems in Europe and other Western nations for various strategic objectives,” the diplomatic service of the European Union wrote in a report last month when the network included only 23 websites.

In recent weeks, the sites have included themes that seem intended to stoke the partisan fires in the United States before November’s presidential election.

Last month, articles appeared on two of Mr. Dougan’s newer fake sites, The Houston Post and The Flagstaff Post, detailing a baseless claim that the F.B.I. had planted an eavesdropping device in Mr. Trump’s office at Mar-a-Lago in Florida.

Some of the new sites have names, like Right Review and Red State Report, that suggest a conservative political bent. In April, a site that researchers also linked to Mr. Dougan offered “major cryptocurrency rewards” for leaks of information about American officials, singling out two prosecutors and a judge involved in the criminal cases against Mr. Trump.

“If the site was mine,” he wrote in response to a question about it, “I would want people to give documents on any dirty politician, Republican, Democrat or other.”